impala-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dan Hecht (Code Review)" <>
Subject [Impala-ASF-CR] IMPALA-4080, IMPALA-3638: Introduce ExecNode::Codegen()
Date Tue, 18 Oct 2016 00:54:59 GMT
Dan Hecht has posted comments on this change.

Change subject: IMPALA-4080, IMPALA-3638: Introduce ExecNode::Codegen()

Patch Set 4:

File be/src/exec/exec-node.h:

PS4, Line 72:  If overridden in subclass, must also
            :   /// call superclass's Codegen().
does it matter whether this happens first or last?
File be/src/exec/

Line 148:   }
why not do this in ExecNode::Prepare() to avoid having to do the same thing in every derived
File be/src/exec/partitioned-hash-join-builder.h:

PS4, Line 88: whether the codegen was
            :   /// enabled 
we only call this if codegen is enabled, right?  So I'm not sure what this part means (maybe
just delete it).
File be/src/exprs/

Line 409:   // TODO: fix this
not for this change, but i think this timestamp stuff may have been fixed by moving things
out of the IR.
File be/src/runtime/runtime-state.h:

PS4, Line 161: 'codegen_'.
public method comments shouldn't really talk about private members. how about "... for this
fragment instance" or something like that.
File fe/src/main/java/org/apache/impala/planner/

Line 532
why is it okay to always enable codegen now, whereas before we were so careful to only enable
it based on this heuristic?
File tests/query_test/

Line 80:     vector.get_value('exec_option')['disable_codegen'] = 1
this doesn't invalidate any of the preexisting test cases?

To view, visit
To unsubscribe, visit

Gerrit-MessageType: comment
Gerrit-Change-Id: I207566bc9f4c6a159271ecdbc4bbdba3d78c6651
Gerrit-PatchSet: 4
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Michael Ho <>
Gerrit-Reviewer: Dan Hecht <>
Gerrit-Reviewer: Marcel Kornacker <>
Gerrit-Reviewer: Michael Ho <>
Gerrit-Reviewer: Tim Armstrong <>
Gerrit-HasComments: Yes

View raw message