Tim Armstrong has posted comments on this change. Change subject: IMPALA-3838: Codegen EvalRuntimeFilters(). ......................................................................
Patch Set 2: (2 comments) http://gerrit.cloudera.org:8080/#/c/4833/1/be/src/exec/hdfs-parquet-scanner-ir.cc File be/src/exec/hdfs-parquet-scanner-ir.cc: Line 73: if (!filter_ctxs_[i]->Eval(row)) { > Actually, RuntimeFilter::Eval() will return true if the filter hasn't arriv I had it reversed - as far as I can see, before filter arrival, it will increment 'considered' and not 'rejected' for every row processed even though the filter is not available yet. I couldn't convince myself that the code was right, so I added some logging and it looks like it's doing something that doesn't make sense: IMPALA-4495 http://gerrit.cloudera.org:8080/#/c/4833/2/be/src/runtime/runtime-filter.h File be/src/runtime/runtime-filter.h: PS2, Line 61: tuple row Is it a tuple row? I though it was a raw value pointer? -- To view, visit http://gerrit.cloudera.org:8080/4833 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I27114869840e268d17e91d6e587ef811628e3837 Gerrit-PatchSet: 2 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Michael Ho <k...@cloudera.com> Gerrit-Reviewer: Michael Ho <k...@cloudera.com> Gerrit-Reviewer: Mostafa Mokhtar <mmokh...@cloudera.com> Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-HasComments: Yes