Qifan Chen has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/16419 )

Change subject: IMPALA-10078: Proper codegen for KuduPartitionExpr
......................................................................


Patch Set 3:

(5 comments)

Thanks a lot for the explanation. Appreciate it.

http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exec/kudu-util-ir.cc
File be/src/exec/kudu-util-ir.cc:

http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exec/kudu-util-ir.cc@32
PS2, Line 32: DCHECK(success);
> I copied these functions from kudu-util.cc into this new file, it already h
My understanding is that DCHECK is applied even in release code. If success is 
false, we will stop here and bail out, right?


http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exec/kudu-util-ir.cc@35
PS2, Line 35: Invalid TimestampValue in function Convert
> You mean the name of this conversion function? I added it to the message.
Yes. The idea is that we can know for sure which function is involved in the 
message.


http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exprs/kudu-partition-expr-ir.cc
File be/src/exprs/kudu-partition-expr-ir.cc:

http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exprs/kudu-partition-expr-ir.cc@40
PS2, Line 40: *row = kudu_row;
            :   *partitioner = kudu_partitioner;
> Why do you think it's better?
Just to make sure these two pointers are always valid to assign values to. 
Maybe they are every time?


http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exprs/kudu-partition-expr.cc
File be/src/exprs/kudu-partition-expr.cc:

http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exprs/kudu-partition-expr.cc@152
PS2, Line 152: lvm::Function* const write_kudu_fn =
             :       codegen->GetFunction(IRFunction::WRITE_KUDU_VALUE, false);
> I don't quite understand it, could you help me?
I mean people changes the declaration of Impala::Status.


http://gerrit.cloudera.org:8080/#/c/16419/2/be/src/exprs/kudu-partition-expr.cc@229
PS2, Line 229: tus KuduPartitionExpr::GetCodegendComputeFnImpl(
             :     LlvmCodeGen* codegen, llvm::Function** fn) {
             :   llvm::Function* const kudu_partition_row_fn =
             :       codegen->GetFunction(IRFunction::GET_KUDU_PARTI
> It's not redundant, 'next_eval_child_block' is assigned to 'current_eval_ch
Okay. Thanks for the explanation.



--
To view, visit http://gerrit.cloudera.org:8080/16419
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ifcae34f71b407837e2c5f1b97aa230e490a268df
Gerrit-Change-Number: 16419
Gerrit-PatchSet: 3
Gerrit-Owner: Daniel Becker <[email protected]>
Gerrit-Reviewer: Csaba Ringhofer <[email protected]>
Gerrit-Reviewer: Daniel Becker <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Qifan Chen <[email protected]>
Gerrit-Reviewer: Thomas Tauber-Marshall <[email protected]>
Gerrit-Reviewer: Tim Armstrong <[email protected]>
Gerrit-Comment-Date: Mon, 14 Sep 2020 14:07:15 +0000
Gerrit-HasComments: Yes

Reply via email to