Tim Armstrong has uploaded a new patch set (#2). Change subject: PREVIEW ONLY: IMPALA-2737: per-partition processing in agg ......................................................................
PREVIEW ONLY: IMPALA-2737: per-partition processing in agg Separates out evaluation of agg input rows from insertion into the hash table. This allows us to do a radix-sort of rows based on hash value and process a partition at a time. This sets things up to do optimisations like prefetching for aggregations and joins. TODO: update comments with changed IR TODO: improve performance - there is too much indirection with the different arrays in ExprValuesBuffer and PartitionedBatch. NOTE: this depends on the LLVM 3.8 patch Change-Id: I93adce55e6d9302dc191d95058440c8a9fa48f48 --- M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/impala-ir.cc M be/src/exec/CMakeLists.txt A be/src/exec/hash-partitioning-ir.cc A be/src/exec/hash-partitioning.cc A be/src/exec/hash-partitioning.h M be/src/exec/hash-table-test.cc M be/src/exec/hash-table.cc M be/src/exec/hash-table.h M be/src/exec/hash-table.inline.h M be/src/exec/partitioned-aggregation-node-ir.cc M be/src/exec/partitioned-aggregation-node.cc M be/src/exec/partitioned-aggregation-node.h M be/src/exec/partitioned-hash-join-node.cc 14 files changed, 969 insertions(+), 349 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala refs/changes/91/2691/2 -- To view, visit http://gerrit.cloudera.org:8080/2691 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: newpatchset Gerrit-Change-Id: I93adce55e6d9302dc191d95058440c8a9fa48f48 Gerrit-PatchSet: 2 Gerrit-Project: Impala Gerrit-Branch: cdh5-trunk Gerrit-Owner: Tim Armstrong <[email protected]>
