RFR: 8309130: x86_64 AVX512 intrinsics for Arrays.sort methods (int, long, float and double arrays) [v39]

[iaroslavski]

On Tue, 19 Sep 2023 21:44:00 GMT, iaroslavski <duke at openjdk.org> wrote:

>> Srinivas Vamsi Parasa has updated the pull request incrementally with two additional commits since the last revision:
>> 
>>  - Update DualPivotQuicksort.java
>>  - Rename arraySort and arrayPartition Java methods to sort and partition. Cleanup some comments
>
> Hi Vamsi,
> 
> The first parameter of introduced method ``sort(Class<?> elemType, A array, long offset, int low, int high, SortOperation<A> so)``
> is ``elemType`` (int.class, long,class etc.). The third parameter is ``offset`` and it depends on ``elemType``. For example,
> if ``elemType`` is int.class, offset is Unsafe.ARRAY_INT_BASE_OFFSET etc.
> 
> Can you detect offset inside intrinsic (C++ code) and remove it from Java code?

> Hello Vladimir (@iaroslavski ),
> 
> Could you please file a separate PR for integrating the `ArraysSort.java` JMH benchmark?
> 
> > 2. You introduced benchmarking class test/micro/org/openjdk/bench/java/util/ArraysSort.java, that's great,
> >    but there is the same class in PR https://raw.githubusercontent.com/openjdk/jdk/42e17e45b1adc4d77ba5549770ce591d96d4b1fe/test/micro/org/openjdk/bench/java/util/ArraysSort.java
> >    which covers all types (not int/long/float/double only) and there are different data inputs (not random only).
> >    Could you please switch to the more powerful ArraysSort class?
> 
> Thanks, Vamsi

Hi Vamsi,

I'm not sure about new PR  for ArraysSort.java. Why do we need separate PR?
You can take my version of ArraysSort.java from https://github.com/openjdk/jdk/pull/13568,
no any objections from my side.

-------------

PR Comment: https://git.openjdk.org/jdk/pull/14227#issuecomment-1727103206