Re: RFR 8154049: DualPivot sorting calculates incorrect runs for nearly sorted arrays

2016-05-03 Thread Paul Sandoz
> On 3 May 2016, at 14:50, Aleksey Shipilev wrote: > > On 04/21/2016 04:24 PM, Paul Sandoz wrote: >> Please review: >> >> http://cr.openjdk.java.net/~psandoz/jdk9/JDK-8154049-dual-pivot-sort-one-off-error/webrev/ > > Ew. It is a complete brain-twister to

Re: RFR 8154049: DualPivot sorting calculates incorrect runs for nearly sorted arrays

2016-05-03 Thread Aleksey Shipilev
On 04/21/2016 04:24 PM, Paul Sandoz wrote: > Please review: > > http://cr.openjdk.java.net/~psandoz/jdk9/JDK-8154049-dual-pivot-sort-one-off-error/webrev/ Ew. It is a complete brain-twister to understand the original code, the original improvement, the bug, and the fix -- all four! This is a

RFR 8154049: DualPivot sorting calculates incorrect runs for nearly sorted arrays

2016-04-21 Thread Paul Sandoz
Hi, Please review: http://cr.openjdk.java.net/~psandoz/jdk9/JDK-8154049-dual-pivot-sort-one-off-error/webrev/ https://bugs.openjdk.java.net/browse/JDK-8154049 A bug was introduced with the fix for: https://bugs.openjdk.java.net/browse/JDK-8080945 Improve the performance of primitive