spark-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From velvia <>
Subject [GitHub] spark issue #15219: [SPARK-14098][SQL] Generate Java code to build CachedCol...
Date Thu, 08 Dec 2016 20:32:33 GMT
Github user velvia commented on the issue:
    Hi, in my experience, interfaces which are implemented by final methods are inlined just
fine in the JVM JIT…. but you might know more about this. 
    The point is more simplification and that the writers and setters are not needed by everyone….
    I would think an API to iterate over non-null values would also be good, but we can’t
use the Iterator API in Java, it has too high overhead….. :-p
    > On Dec 8, 2016, at 12:29 PM, Kazuaki Ishizaki <> wrote:
    > @velvia <> Thank you for your comment. It is interesting
from the view of simplification.
    > On the other hand, I think that current implementation thinks about performance to
encourage method inlining for getter/setter (e.g. getInt()) and avoid interface call in JIT
    > Let me think about this among simplification and performance. This refactoring should
be addressed by another PR.
    > —
    > You are receiving this because you were mentioned.
    > Reply to this email directly, view it on GitHub <>,
or mute the thread <>.

If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at or file a JIRA ticket
with INFRA.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message