[
https://issues.apache.org/jira/browse/DRILL-3912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14986640#comment-14986640
]
ASF GitHub Bot commented on DRILL-3912:
---------------------------------------
Github user jinfengni commented on a diff in the pull request:
https://github.com/apache/drill/pull/189#discussion_r43714802
--- Diff:
exec/java-exec/src/main/java/org/apache/drill/exec/expr/EvaluationVisitor.java
---
@@ -106,19 +177,30 @@ public HoldingContainer
visitFunctionCall(FunctionCall call, ClassGenerator<?> g
@Override
public HoldingContainer visitBooleanOperator(BooleanOperator op,
ClassGenerator<?> generator) throws RuntimeException {
+ HoldingContainer hc = getPrevious(op, generator.getMappingSet());
--- End diff --
You put the check whether it's common subexpression in EvalVisitor, when
handling with different expression.
I feel the code would be much cleaner, if we add a CSEFilter in a similar
to ConstantFilter. The logic will go through ConstantFilter -> CSEFilter ->
EvalVisitor. The check of common subexpression and adding to the Hashmap would
be done solely in CSEFilter:visitiUnknow(). In AbstractExrVisitor, other
methods will by default come to visitUnknow(). In this way, there is no need
to repeat the checking of common subexpression for different expression in
EvalVisitor; actually no change has to be done to EvalVisitor.
> Common subexpression elimination in code generation
> ---------------------------------------------------
>
> Key: DRILL-3912
> URL: https://issues.apache.org/jira/browse/DRILL-3912
> Project: Apache Drill
> Issue Type: Improvement
> Reporter: Steven Phillips
> Assignee: Jinfeng Ni
>
> Drill currently will evaluate the full expression tree, even if there are
> redundant subtrees. Many of these redundant evaluations can be eliminated by
> reusing the results from previously evaluated expression trees.
> For example,
> {code}
> select a + 1, (a + 1)* (a - 1) from t
> {code}
> Will compute the entire (a + 1) expression twice. With CSE, it will only be
> evaluated once.
> The benefit will be reducing the work done when evaluating expressions, as
> well as reducing the amount of code that is generated, which could also lead
> to better JIT optimization.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
