Tim Armstrong has posted comments on this change.

Change subject: IMPALA-1583: Simplify 
PartitionedHashJoinNode::ProcessProbeBatch()
......................................................................


Patch Set 1:

(2 comments)

Did you do any perf testing? It shouldn't make a difference but there's always 
a chance that we somehow caused the compiler to produce different output.

I had some questions about taking this further and completing the 
simplification. It would be nice to do it all in one go but I'm ok with this 
patch as-is.

http://gerrit.cloudera.org:8080/#/c/2893/1/be/src/exec/partitioned-hash-join-node-ir.cc
File be/src/exec/partitioned-hash-join-node-ir.cc:

Line 285: next_row:
How hard is it to restructure to avoid the goto? E.g. something like

  if (no probe row) {
    get next probe row
  }
  process probe row


Line 298:     current_probe_row_ = probe_batch_iterator.Get();
Can we factor the probe row establishment out into its own function too? Might 
make it easier to see the high-level algorithm.


-- 
To view, visit http://gerrit.cloudera.org:8080/2893
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Ie2091bdf97ab34c5cdc84e84394c579a5b36afc0
Gerrit-PatchSet: 1
Gerrit-Project: Impala
Gerrit-Branch: cdh5-trunk
Gerrit-Owner: Michael Ho <[email protected]>
Gerrit-Reviewer: Michael Ho <[email protected]>
Gerrit-Reviewer: Tim Armstrong <[email protected]>
Gerrit-HasComments: Yes

Reply via email to