GitHub user hvanhovell opened a pull request:
https://github.com/apache/spark/pull/20854
[SPARK-23712][SQL] Interpreted UnsafeRowJoiner [WIP]
## What changes were proposed in this pull request?
This PR adds an interpreted version of `UnsafeRowJoiner` to Spark SQL.
Its performance is almost to par with the code generated `UnsafeRowJoiner`.
There seems to be an overhead of 10ns per call. It might be an idea to not use
code generation at all for an `UnsafeRowJoiner`
## How was this patch tested?
Modified existing row joiner tests.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/hvanhovell/spark SPARK-23712
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/20854.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #20854
----
commit b637ded5ddd38f58e2c0d1b5172ebed5cb9014e2
Author: Herman van Hovell <hvanhovell@...>
Date: 2018-03-17T13:42:13Z
Add interpreted unsafe row joiner
commit d0b40a9ff6368051d737224dd9931a7ef1b428cb
Author: Herman van Hovell <hvanhovell@...>
Date: 2018-03-18T12:16:30Z
Add benchmark
----
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
