Github user vanzin commented on the issue:

    https://github.com/apache/spark/pull/22114
  
    FYI I wrote a small micro bench to pound on the pool in both the old way 
(mixed on heap and off heap allocations) and new way (only off heap) and it 
actually seems faster if all allocations use the `buffer()` call. So I guess 
we're safe here.


---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to