[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Internal Jenkins has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 10: Verified+1 -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 10 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Internal Jenkins has submitted this change and it was merged. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and interpreted paths. A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. Instead in this patch constant args are evaluated once and cached. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into an automatic buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Reviewed-on: http://gerrit.cloudera.org:8080/4838 Reviewed-by: Tim Armstrong Tested-by: Internal Jenkins --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/aggregate-functions-ir.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/exprs/utility-functions-ir.cc M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/sr
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 10: Hit IMPALA-3040 -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 10 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Internal Jenkins has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 10: Verified-1 Build failed: http://sandbox.jenkins.cloudera.com/job/impala-external-gerrit-verify-merge-ASF/435/ -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 10 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 10: Code-Review+2 (1 comment) Carry +2 http://gerrit.cloudera.org:8080/#/c/4838/8/be/src/exprs/aggregate-functions-ir.cc File be/src/exprs/aggregate-functions-ir.cc: Line 1030: // significant bits overlap in memory. > I guess these should really have the LITTLE_ENDIAN checks too, right? Reworded the comment to be less little-endian specific. I think if we were to port this to big-endian, we'd add padding to DecimalVal so that the least significant bits lined up (which means the field offsets would be different). I don't think this code would have to change. -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 10 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Hello Internal Jenkins, Dan Hecht, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/4838 to look at the new patch set (#9). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and interpreted paths. A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. Instead in this patch constant args are evaluated once and cached. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into an automatic buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/aggregate-functions-ir.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/exprs/utility-functions-ir.cc M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/udf
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Dan Hecht has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 8: Code-Review+2 (1 comment) http://gerrit.cloudera.org:8080/#/c/4838/8/be/src/exprs/aggregate-functions-ir.cc File be/src/exprs/aggregate-functions-ir.cc: Line 1030: // of the val16 should be initialized to all 0's. I guess these should really have the LITTLE_ENDIAN checks too, right? -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 8 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 8: Ran into a bug in the patch where histogram was interpreting a val4 or val8 as a val16, and seeing uninitialised bits. This seems to be an expected way to use DecimalVal in the UDF interface. The fix is to memset() the whole value to 0 when allocating it. Documented this behaviour a bit better. -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 8 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Hello Internal Jenkins, Dan Hecht, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/4838 to look at the new patch set (#8). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and interpreted paths. A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. Instead in this patch constant args are evaluated once and cached. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into an automatic buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/aggregate-functions-ir.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/exprs/utility-functions-ir.cc M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/udf
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Hello Internal Jenkins, Dan Hecht, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/4838 to look at the new patch set (#7). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and interpreted paths. A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. Instead in this patch constant args are evaluated once and cached. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into an automatic buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/aggregate-functions-ir.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/exprs/utility-functions-ir.cc M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/udf
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Internal Jenkins has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 6: Verified-1 Build failed: http://sandbox.jenkins.cloudera.com/job/impala-external-gerrit-verify-merge-ASF/430/ -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 6 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Internal Jenkins Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 6: Code-Review+2 rebase -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 6 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Dan Hecht has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 5: Code-Review+2 -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 5 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 4: (11 comments) http://gerrit.cloudera.org:8080/#/c/4838/4//COMMIT_MSG Commit Message: Line 18: arguments for ScalarFnCall expressions on both the codegen'd and > missing words Done PS4, Line 36: temporary > automatic? at least I would have understood this a little quicker with tha Done http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/agg-fn-evaluator.cc File be/src/exprs/agg-fn-evaluator.cc: PS4, Line 158: ", > is it worth differentiating these messages to make debugging easier? Can't hurt. http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/anyval-util.h File be/src/exprs/anyval-util.h: PS4, Line 297: The returned AnyVal is not initialized. > this sentence is repeated Done http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/expr-context.h File be/src/exprs/expr-context.h: PS4, Line 66: fragment > do you mean fragment instance? It's kind-of unclear what the direction here is. There's a FRAGMENT_LOCAL concept in the udf interface that this corresponds to, so I think this is consistent at the moment (although maybe it should be FRAGMENT_INSTANCE_LOCAL everywhere). http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/scalar-fn-call.cc File be/src/exprs/scalar-fn-call.cc: PS4, Line 188: fragment > fragment or fragment instance? Reworded to avoid making the distinction and be clearer. It's not clear that the FRAGMENT_LOCAL terminology is accurate, but I think that's out of scope for this fix. PS4, Line 189: cases > not sure what this sentence is saying. what cases? Reworded to be clearer. It's copied when the ExprContext is cloned. Line 205: // children. > do we document the calling convention somewhere? I think a brief explanatio Updated the comment. The UDF calling convention is in udf/udf.h and the contents of the buffers are in udf/udf-internal.h PS4, Line 340: optimise out the buffer > is this also enabling constant propagation for constant args, or was that a The constant was propagated but the side-effect of storing to the varargs buffer couldn't be optimised out. PS4, Line 370: Either no varargs or arguments before varargs begin > this comment seems unnecessary with the new conditional Done http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/runtime/free-pool.h File be/src/runtime/free-pool.h: Line 73: size = std::max(8, size); > seems fine, but why is it required? Not required but it's pointless making smaller allocations, since it's aligned to 8 bytes by the MemPool anyway. Added a comment -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 4 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has uploaded a new patch set (#5). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and interpreted paths. A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. Instead in this patch constant args are evaluated once and cached. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into an automatic buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/util/bit-util.h M testdata/workloads/functional-query/queries/QueryTest/exprs.test M testdata/workloads/functional-query/queries/QueryTest/uda.test A testdata/workloads/targeted-perf/queries/primiti
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Dan Hecht has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 4: (11 comments) Looks good. I think just comments about comments. http://gerrit.cloudera.org:8080/#/c/4838/4//COMMIT_MSG Commit Message: Line 18: arguments for ScalarFnCall expressions on both the codegen'd and missing words PS4, Line 36: temporary automatic? at least I would have understood this a little quicker with that terminology. http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/agg-fn-evaluator.cc File be/src/exprs/agg-fn-evaluator.cc: PS4, Line 158: ", is it worth differentiating these messages to make debugging easier? http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/anyval-util.h File be/src/exprs/anyval-util.h: PS4, Line 297: The returned AnyVal is not initialized. this sentence is repeated http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/expr-context.h File be/src/exprs/expr-context.h: PS4, Line 66: fragment do you mean fragment instance? http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/exprs/scalar-fn-call.cc File be/src/exprs/scalar-fn-call.cc: PS4, Line 188: fragment fragment or fragment instance? PS4, Line 189: cases not sure what this sentence is saying. what cases? Line 205: // children. do we document the calling convention somewhere? I think a brief explanation would help casual readers understand why the interpreted path and codegen path set things up the way they do. PS4, Line 340: optimise out the buffer is this also enabling constant propagation for constant args, or was that already happening? PS4, Line 370: Either no varargs or arguments before varargs begin this comment seems unnecessary with the new conditional http://gerrit.cloudera.org:8080/#/c/4838/4/be/src/runtime/free-pool.h File be/src/runtime/free-pool.h: Line 73: size = std::max(8, size); seems fine, but why is it required? -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 4 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Dan Hecht Gerrit-Reviewer: Michael Ho Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has posted comments on this change. Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. Patch Set 3: (3 comments) http://gerrit.cloudera.org:8080/#/c/4838/3/be/src/codegen/llvm-codegen.h File be/src/codegen/llvm-codegen.h: Line 396: /// Initialise a constant global string and returns a pointer to it. Need to doc that it's an i8* http://gerrit.cloudera.org:8080/#/c/4838/3/be/src/exprs/anyval-util.h File be/src/exprs/anyval-util.h: Line 21: #include "gutil/strings/substitute.h" Move to .cc http://gerrit.cloudera.org:8080/#/c/4838/3/be/src/exprs/scalar-fn-call.cc File be/src/exprs/scalar-fn-call.cc: Line 337: // Get IR i8* pointer to a varargs buffer. We allocate the buffer with Alloca so that needs updating, not an i8 -- To view, visit http://gerrit.cloudera.org:8080/4838 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 Gerrit-PatchSet: 3 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Tim Armstrong Gerrit-Reviewer: Tim Armstrong Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has uploaded a new patch set (#4). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. This is improved by separating out constant and non-constant args and only evaluating the non-constant args. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into a temporary buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/util/bit-util.h M testdata/workloads/functional-query/queries/QueryTest/exprs.test M testdata/workloads/functional-query/queries/QueryTest/uda.test A testdata/workloads/targete
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has uploaded a new patch set (#3). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. This is improved by separating out constant and non-constant args and only evaluating the non-constant args. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into a temporary buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.cc M be/src/exprs/scalar-fn-call.h M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/util/bit-util.h M testdata/workloads/functional-query/queries/QueryTest/exprs.test M testdata/workloads/functional-query/queries/QueryTest/uda.test A testdata/workloads/targete
[Impala-ASF-CR] IMPALA-4302,IMPALA-2379: constant expr arg fixes
Tim Armstrong has uploaded a new patch set (#2). Change subject: IMPALA-4302,IMPALA-2379: constant expr arg fixes .. IMPALA-4302,IMPALA-2379: constant expr arg fixes This patch fixes two issues around handling of constant expr args. The patches are combined because they touch some of the same code and depend on some of the same memory management cleanup. First, it fixes IMPALA-2379, where constant expr args were not visible to UDAFs. The issue is that the input exprs need to be opened before calling the UDAF Init() function. Second, it avoids overhead from repeated evaluation of constant arguments for ScalarFnCall expressions on both the codegen'd and A common example is an IN predicate with a long list of constant values. The interpreted path was inefficient because it always evaluated all children expressions. This is improved by separating out constant and non-constant args and only evaluating the non-constant args. The memory management of the AnyVal* objects was somewhat nebulous - adjusted it so that they're allocated from ExprContext::mem_pool_, which has the correct lifetime. The codegen'd path was inefficient only with varargs - with fixed arguments the LLVM optimiser is able to infer after inlining that the expressions are constant and remove all evaluation. However, for varargs it stores the vararg values into a heap-allocated buffer. The LLVM optimiser is unable to remove these stores because they have a side-effect that is visible to code outside the function. The codegen'd path is improved by evaluating varargs into a temporary buffer that can be optimised out. We also make a small related change to bake the string constants into the codegen'd code. Testing: Ran exhaustive build. Added regression test for IMPALA-2379 and MemPool test for aligned allocation. Added a test for in predicates with constant strings. Perf: Added a targeted query that demonstrates the improvement. Also manually validated the non-codegend perf. Also ran TPC-H and targeted perf queries locally - didn't see any significant changes. ++---+---++-++---++-+---+ | Workload | Query | File Format | Avg(s) | Base Avg(s) | Delta(Avg) | StdDev(%) | Base StdDev(%) | Num Clients | Iters | ++---+---++-++---++-+---+ | TARGETED-PERF(_20) | primitive_filter_in_predicate | parquet / none / none | 1.19 | 9.82| I -87.85% | 3.82% | 0.71%| 1 | 10| ++---+---++-++---++-+---+ (I) Improvement: TARGETED-PERF(_20) primitive_filter_in_predicate [parquet / none / none] (9.82s -> 1.19s [-87.85%]) +--++--+--++---+--+--++++---+ | Operator | % of Query | Avg | Base Avg | Delta(Avg) | StdDev(%) | Max | Base Max | Delta(Max) | #Hosts | #Rows | Est #Rows | +--++--+--++---+--+--++++---+ | 01:AGGREGATE | 14.39% | 155.88ms | 214.61ms | -27.37%| 2.68% | 163.38ms | 227.53ms | -28.19%| 1 | 1 | 1 | | 00:SCAN HDFS | 85.60% | 927.46ms | 9.43s| -90.16%| 4.49% | 1.01s| 9.50s| -89.42%| 1 | 13.77K | 14.05K| +--++--+--++---+--+--++++---+ Change-Id: I45c3ed8c9d7a61e94a9b9d6c316e8a53d9ff6c24 --- M be/src/benchmarks/in-predicate-benchmark.cc M be/src/codegen/gen_ir_descriptions.py M be/src/codegen/llvm-codegen.cc M be/src/codegen/llvm-codegen.h M be/src/exec/partitioned-aggregation-node.cc M be/src/exprs/agg-fn-evaluator.cc M be/src/exprs/anyval-util.cc M be/src/exprs/anyval-util.h M be/src/exprs/case-expr.cc M be/src/exprs/expr-context.h M be/src/exprs/expr.cc M be/src/exprs/expr.h M be/src/exprs/hive-udf-call.cc M be/src/exprs/in-predicate.h M be/src/exprs/literal.cc M be/src/exprs/scalar-fn-call.h M be/src/runtime/free-pool-test.cc M be/src/runtime/free-pool.h M be/src/runtime/mem-pool-test.cc M be/src/runtime/mem-pool.cc M be/src/runtime/mem-pool.h M be/src/testutil/test-udas.cc M be/src/udf/udf-internal.h M be/src/udf/udf-test-harness.cc M be/src/udf/udf.cc M be/src/util/bit-util.h M testdata/workloads/functional-query/queries/QueryTest/exprs.test M testdata/workloads/functional-query/queries/QueryTest/uda.test A testdata/workloads/targeted-perf/queries/primitive_filter_i
