[
https://issues.apache.org/jira/browse/LUCENE-9918?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17391245#comment-17391245
]
Gautam Worah edited comment on LUCENE-9918 at 8/1/21, 10:13 PM:
----------------------------------------------------------------
Hey [~gsmiller] , I noticed that in the micro benchmark code in your
lucene-pfor-benchmark [repo |#L15], the main loop runs 10 times I think?
Some
[sources|http://daniel-strecker.com/blog/2020-01-14_auto_vectorization_in_java/#Output%20Interpretation]
suggest that usually the JIT compiler compiles and optimizes statements as and
when it sees that a particular operation is repeated multiple times. So it
first optimizes them a little and them some more iff it sees them again. So
maybe we just need to repeat the experiment with say 100k iterations (and check
if it is auto vectorized)?
was (Author: gworah):
Hey [~gsmiller] , I noticed that in the micro benchmark code in your
lucene-pfor-benchmark [repo |#L15],] the main loop runs 10 times I think?
Some
[sources|http://daniel-strecker.com/blog/2020-01-14_auto_vectorization_in_java/#Output%20Interpretation]
suggest that usually the JIT compiler compiles and optimizes statements as and
when it sees that a particular operation is repeated multiple times. So it
first optimizes them a little and them some more iff it sees them again. So
maybe we just need to repeat the experiment with say 100k iterations (and check
if it is auto vectorized)?
> Can PForUtil be further auto-vectorized?
> ----------------------------------------
>
> Key: LUCENE-9918
> URL: https://issues.apache.org/jira/browse/LUCENE-9918
> Project: Lucene - Core
> Issue Type: Task
> Components: core/codecs
> Affects Versions: main (9.0)
> Reporter: Greg Miller
> Priority: Minor
>
> While working on LUCENE-9850, we discovered the loop in PForUtil::prefixSumOf
> is not getting auto-vectorized by the HotSpot compiler. We tried a few
> different tweaks to see if we could change this, but came up empty. There are
> some additional suggestions in the related
> [PR|https://github.com/apache/lucene/pull/69#discussion_r608412309] that
> could still be experimented with, and it may be worth doing so to see if
> further improvements could be squeezed out.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
