feat: plan IN/EXISTS subquery decorrelation as RightSemi by default by Dandandan · Pull Request #21809 · apache/datafusion

Dandandan · 2026-04-23T19:01:28Z

Which issue does this PR close?

Closes #.

Rationale for this change

Are these changes tested?

Yes — existing tests are updated to the new plan shape:

cargo test -p datafusion-optimizer (unit + integration snapshots regenerated via cargo insta)
cargo test --test sqllogictests -p datafusion-sqllogictest (all 463 SLT files)
INCLUDE_TPCH=true cargo test --test sqllogictests -p datafusion-sqllogictest (all 464 SLT files, including TPC-H plan snapshots)
cargo fmt --all / cargo clippy -p datafusion-optimizer --all-targets -- -D warnings

Are there any user-facing changes?

The optimized logical/physical plan for IN/EXISTS subqueries now shows RightSemi with the subquery as the left child. Query results are unchanged.

Change DecorrelatePredicateSubquery to emit RightSemi (with the subquery on the left and the outer query on the right) instead of LeftSemi for non-negated IN/EXISTS predicates. Semantics are preserved: RightSemi returns rows from the right input (the outer query) that have a match on the left, which matches the previous LeftSemi output. The NOT IN / NOT EXISTS path still uses LeftAnti (including null-aware handling) and is unchanged. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Dandandan · 2026-04-23T19:02:19Z

run benchmark tpch tpch10 tpcds

adriangbot · 2026-04-23T19:05:00Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4307064194-1783-l6p5m 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing right_semi_default (265fe55) to 067ba4b (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-23T19:05:48Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4307064194-1784-sgf6j 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing right_semi_default (265fe55) to 067ba4b (merge-base) diff using: tpch10
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-23T19:05:50Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4307064194-1785-sfrtv 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing right_semi_default (265fe55) to 067ba4b (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-23T19:17:50Z

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highmem-16 (12 vCPU / 65 GiB)

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Details

Comparing HEAD and right_semi_default
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃             right_semi_default ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │ 40.02 / 40.93 ±0.85 / 42.39 ms │ 40.09 / 41.52 ±1.24 / 43.39 ms │ no change │
│ QQuery 2  │ 20.78 / 20.95 ±0.14 / 21.19 ms │ 20.60 / 20.86 ±0.18 / 21.15 ms │ no change │
│ QQuery 3  │ 40.50 / 41.20 ±0.44 / 41.62 ms │ 38.51 / 40.66 ±1.43 / 42.29 ms │ no change │
│ QQuery 4  │ 17.90 / 18.18 ±0.17 / 18.44 ms │ 18.01 / 18.32 ±0.43 / 19.17 ms │ no change │
│ QQuery 5  │ 48.42 / 49.85 ±1.03 / 51.11 ms │ 50.87 / 51.68 ±0.75 / 53.04 ms │ no change │
│ QQuery 6  │ 17.23 / 17.50 ±0.23 / 17.84 ms │ 17.20 / 17.38 ±0.13 / 17.57 ms │ no change │
│ QQuery 7  │ 54.81 / 55.49 ±0.57 / 56.24 ms │ 52.67 / 53.76 ±0.98 / 55.32 ms │ no change │
│ QQuery 8  │ 47.54 / 47.97 ±0.32 / 48.39 ms │ 47.85 / 49.08 ±1.65 / 52.21 ms │ no change │
│ QQuery 9  │ 53.56 / 54.31 ±0.84 / 55.85 ms │ 53.25 / 53.73 ±0.49 / 54.58 ms │ no change │
│ QQuery 10 │ 65.03 / 65.74 ±0.93 / 67.56 ms │ 65.01 / 66.47 ±1.06 / 67.87 ms │ no change │
│ QQuery 11 │ 14.08 / 14.51 ±0.57 / 15.61 ms │ 14.08 / 14.59 ±0.64 / 15.84 ms │ no change │
│ QQuery 12 │ 27.41 / 28.19 ±0.84 / 29.80 ms │ 27.09 / 27.48 ±0.24 / 27.75 ms │ no change │
│ QQuery 13 │ 36.54 / 39.16 ±3.42 / 45.79 ms │ 37.68 / 38.25 ±0.47 / 38.83 ms │ no change │
│ QQuery 14 │ 27.56 / 27.81 ±0.15 / 27.99 ms │ 27.85 / 28.22 ±0.61 / 29.43 ms │ no change │
│ QQuery 15 │ 33.50 / 34.17 ±0.66 / 35.38 ms │ 33.38 / 34.29 ±0.98 / 35.76 ms │ no change │
│ QQuery 16 │ 15.13 / 15.23 ±0.14 / 15.50 ms │ 15.01 / 15.22 ±0.12 / 15.34 ms │ no change │
│ QQuery 17 │ 77.35 / 78.01 ±0.77 / 79.49 ms │ 76.55 / 78.00 ±0.92 / 79.32 ms │ no change │
│ QQuery 18 │ 74.95 / 75.44 ±0.54 / 76.48 ms │ 74.28 / 75.23 ±0.74 / 76.25 ms │ no change │
│ QQuery 19 │ 37.26 / 38.02 ±1.08 / 40.16 ms │ 36.93 / 37.13 ±0.33 / 37.78 ms │ no change │
│ QQuery 20 │ 39.17 / 39.75 ±0.70 / 41.04 ms │ 38.96 / 39.80 ±1.43 / 42.64 ms │ no change │
│ QQuery 21 │ 61.88 / 63.36 ±0.95 / 64.47 ms │ 62.88 / 63.78 ±0.82 / 65.14 ms │ no change │
│ QQuery 22 │ 16.97 / 17.03 ±0.06 / 17.10 ms │ 16.89 / 17.12 ±0.14 / 17.26 ms │ no change │
└───────────┴────────────────────────────────┴────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                 ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                 │ 882.77ms │
│ Total Time (right_semi_default)   │ 882.54ms │
│ Average Time (HEAD)               │  40.13ms │
│ Average Time (right_semi_default) │  40.12ms │
│ Queries Faster                    │        0 │
│ Queries Slower                    │        0 │
│ Queries with No Change            │       22 │
│ Queries with Failure              │        0 │
└───────────────────────────────────┴──────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	5.0s
Peak memory	5.5 GiB
Avg memory	4.8 GiB
CPU user	33.3s
CPU sys	2.4s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	5.0s
Peak memory	5.3 GiB
Avg memory	4.8 GiB
CPU user	33.4s
CPU sys	2.4s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-04-23T19:19:39Z

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highmem-16 (12 vCPU / 65 GiB)

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Details

Comparing HEAD and right_semi_default
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃                 right_semi_default ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │  326.74 / 328.81 ±1.43 / 330.70 ms │  327.89 / 330.43 ±1.90 / 333.18 ms │ no change │
│ QQuery 2  │  136.21 / 139.00 ±1.70 / 140.93 ms │  134.53 / 139.33 ±2.49 / 141.76 ms │ no change │
│ QQuery 3  │  296.67 / 301.21 ±2.42 / 303.75 ms │  297.37 / 300.26 ±2.75 / 304.93 ms │ no change │
│ QQuery 4  │  157.42 / 161.04 ±2.04 / 162.86 ms │  156.14 / 159.45 ±3.12 / 165.25 ms │ no change │
│ QQuery 5  │  440.70 / 445.70 ±4.11 / 452.84 ms │  430.47 / 440.21 ±7.20 / 451.27 ms │ no change │
│ QQuery 6  │  134.45 / 136.29 ±1.79 / 139.65 ms │  134.06 / 135.28 ±1.22 / 137.29 ms │ no change │
│ QQuery 7  │  549.59 / 558.13 ±5.61 / 567.28 ms │  547.34 / 552.20 ±4.07 / 556.76 ms │ no change │
│ QQuery 8  │  466.41 / 471.44 ±4.09 / 475.39 ms │  468.15 / 474.77 ±4.76 / 481.31 ms │ no change │
│ QQuery 9  │  654.77 / 660.69 ±8.04 / 676.61 ms │ 662.47 / 677.45 ±11.99 / 688.71 ms │ no change │
│ QQuery 10 │  332.35 / 339.02 ±5.54 / 344.96 ms │  330.74 / 341.66 ±7.64 / 349.68 ms │ no change │
│ QQuery 11 │  106.93 / 111.28 ±4.22 / 119.31 ms │  104.43 / 109.62 ±4.62 / 117.66 ms │ no change │
│ QQuery 12 │  205.05 / 207.29 ±2.86 / 212.86 ms │  202.17 / 209.31 ±7.08 / 220.30 ms │ no change │
│ QQuery 13 │  311.33 / 317.18 ±6.49 / 329.15 ms │  313.85 / 318.64 ±2.89 / 322.14 ms │ no change │
│ QQuery 14 │  183.28 / 186.18 ±3.42 / 192.54 ms │  185.24 / 186.22 ±1.20 / 188.53 ms │ no change │
│ QQuery 15 │  333.64 / 338.74 ±5.84 / 349.46 ms │  332.50 / 334.66 ±1.99 / 337.20 ms │ no change │
│ QQuery 16 │     79.75 / 82.23 ±1.94 / 84.78 ms │     80.51 / 81.60 ±1.44 / 84.37 ms │ no change │
│ QQuery 17 │  774.37 / 779.59 ±3.39 / 785.05 ms │  763.74 / 769.55 ±6.40 / 780.87 ms │ no change │
│ QQuery 18 │ 784.62 / 798.38 ±10.46 / 811.43 ms │ 779.70 / 791.63 ±10.75 / 807.28 ms │ no change │
│ QQuery 19 │ 270.64 / 281.67 ±12.96 / 303.97 ms │ 269.45 / 287.32 ±15.58 / 304.92 ms │ no change │
│ QQuery 20 │  311.01 / 315.84 ±3.19 / 320.04 ms │  307.50 / 312.35 ±3.82 / 318.86 ms │ no change │
│ QQuery 21 │  823.74 / 830.03 ±4.30 / 837.27 ms │  818.67 / 826.94 ±4.57 / 832.74 ms │ no change │
│ QQuery 22 │     83.27 / 86.56 ±1.79 / 88.63 ms │     83.43 / 85.14 ±1.53 / 87.44 ms │ no change │
└───────────┴────────────────────────────────────┴────────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                 ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                 │ 7876.30ms │
│ Total Time (right_semi_default)   │ 7864.02ms │
│ Average Time (HEAD)               │  358.01ms │
│ Average Time (right_semi_default) │  357.46ms │
│ Queries Faster                    │         0 │
│ Queries Slower                    │         0 │
│ Queries with No Change            │        22 │
│ Queries with Failure              │         0 │
└───────────────────────────────────┴───────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	40.0s
Peak memory	12.3 GiB
Avg memory	8.4 GiB
CPU user	423.5s
CPU sys	20.9s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	40.0s
Peak memory	12.4 GiB
Avg memory	8.4 GiB
CPU user	421.9s
CPU sys	21.1s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-04-23T19:22:46Z

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highmem-16 (12 vCPU / 65 GiB)

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Details

Comparing HEAD and right_semi_default
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃                        right_semi_default ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.91 / 7.40 ±0.84 / 9.07 ms │               6.64 / 7.21 ±0.85 / 8.89 ms │     no change │
│ QQuery 2  │        145.28 / 145.83 ±0.46 / 146.67 ms │         143.47 / 143.97 ±0.34 / 144.30 ms │     no change │
│ QQuery 3  │        114.98 / 116.36 ±1.18 / 118.54 ms │         112.33 / 113.00 ±0.50 / 113.82 ms │     no change │
│ QQuery 4  │    1246.27 / 1254.23 ±12.89 / 1279.95 ms │     1230.53 / 1249.72 ±18.63 / 1276.55 ms │     no change │
│ QQuery 5  │        172.02 / 172.70 ±0.46 / 173.28 ms │         170.98 / 172.90 ±1.09 / 174.12 ms │     no change │
│ QQuery 6  │       818.51 / 848.89 ±27.43 / 900.52 ms │        789.62 / 837.84 ±30.59 / 868.20 ms │     no change │
│ QQuery 7  │        332.11 / 334.59 ±2.20 / 338.61 ms │         328.73 / 331.42 ±1.62 / 333.59 ms │     no change │
│ QQuery 8  │        112.90 / 113.38 ±0.25 / 113.57 ms │         112.64 / 113.35 ±0.91 / 115.15 ms │     no change │
│ QQuery 9  │        99.65 / 114.37 ±12.22 / 129.05 ms │          96.03 / 101.91 ±3.58 / 105.69 ms │ +1.12x faster │
│ QQuery 10 │        100.85 / 101.67 ±0.54 / 102.35 ms │         100.66 / 101.02 ±0.31 / 101.39 ms │     no change │
│ QQuery 11 │       854.57 / 874.22 ±21.47 / 910.35 ms │        870.88 / 885.50 ±12.12 / 906.16 ms │     no change │
│ QQuery 12 │           42.90 / 42.98 ±0.09 / 43.13 ms │            43.00 / 43.26 ±0.30 / 43.80 ms │     no change │
│ QQuery 13 │        385.90 / 388.40 ±2.74 / 392.56 ms │         384.99 / 389.44 ±3.26 / 394.42 ms │     no change │
│ QQuery 14 │        967.51 / 977.87 ±5.88 / 984.48 ms │         969.98 / 975.55 ±3.97 / 981.30 ms │     no change │
│ QQuery 15 │           14.51 / 14.79 ±0.31 / 15.31 ms │            14.13 / 14.65 ±0.33 / 15.09 ms │     no change │
│ QQuery 16 │              7.17 / 7.30 ±0.20 / 7.70 ms │               7.08 / 7.21 ±0.12 / 7.44 ms │     no change │
│ QQuery 17 │        219.22 / 220.72 ±1.06 / 222.40 ms │         221.00 / 221.55 ±0.65 / 222.55 ms │     no change │
│ QQuery 18 │        124.56 / 125.22 ±0.57 / 126.08 ms │         123.74 / 124.79 ±0.54 / 125.22 ms │     no change │
│ QQuery 19 │        152.32 / 154.12 ±1.93 / 157.45 ms │         151.91 / 152.73 ±0.68 / 153.76 ms │     no change │
│ QQuery 20 │           12.56 / 13.04 ±0.28 / 13.41 ms │            12.31 / 12.74 ±0.28 / 13.10 ms │     no change │
│ QQuery 21 │           18.73 / 18.96 ±0.18 / 19.21 ms │            18.50 / 19.09 ±0.40 / 19.65 ms │     no change │
│ QQuery 22 │        471.87 / 475.83 ±2.54 / 478.31 ms │         475.44 / 477.12 ±1.71 / 480.29 ms │     no change │
│ QQuery 23 │        801.96 / 806.46 ±5.07 / 816.26 ms │        794.79 / 807.53 ±10.74 / 825.72 ms │     no change │
│ QQuery 24 │        368.84 / 371.03 ±2.21 / 375.20 ms │         368.30 / 370.42 ±2.36 / 374.99 ms │     no change │
│ QQuery 25 │        329.33 / 333.06 ±2.27 / 335.57 ms │         329.27 / 330.86 ±1.14 / 332.55 ms │     no change │
│ QQuery 26 │           76.53 / 77.11 ±0.56 / 78.10 ms │            76.33 / 77.19 ±1.08 / 79.32 ms │     no change │
│ QQuery 27 │              6.81 / 6.93 ±0.15 / 7.21 ms │               6.80 / 6.91 ±0.17 / 7.25 ms │     no change │
│ QQuery 28 │        147.23 / 148.77 ±2.42 / 153.59 ms │         147.35 / 148.81 ±1.37 / 151.03 ms │     no change │
│ QQuery 29 │        269.65 / 272.84 ±3.00 / 278.47 ms │         270.49 / 271.22 ±0.82 / 272.35 ms │     no change │
│ QQuery 30 │           40.46 / 40.98 ±0.34 / 41.46 ms │            40.23 / 40.79 ±0.40 / 41.45 ms │     no change │
│ QQuery 31 │        162.79 / 165.15 ±1.50 / 167.23 ms │         162.51 / 163.65 ±1.19 / 165.81 ms │     no change │
│ QQuery 32 │           12.88 / 13.19 ±0.26 / 13.66 ms │            12.87 / 13.07 ±0.30 / 13.66 ms │     no change │
│ QQuery 33 │        137.79 / 139.30 ±0.86 / 140.23 ms │         137.37 / 138.62 ±0.91 / 139.74 ms │     no change │
│ QQuery 34 │              6.65 / 6.82 ±0.23 / 7.28 ms │               6.66 / 6.85 ±0.25 / 7.35 ms │     no change │
│ QQuery 35 │          98.97 / 99.50 ±0.41 / 100.04 ms │           98.30 / 99.96 ±0.85 / 100.60 ms │     no change │
│ QQuery 36 │              6.38 / 6.64 ±0.20 / 6.91 ms │               6.39 / 6.66 ±0.20 / 6.93 ms │     no change │
│ QQuery 37 │              8.09 / 8.19 ±0.14 / 8.48 ms │               8.02 / 8.13 ±0.09 / 8.26 ms │     no change │
│ QQuery 38 │           84.93 / 87.92 ±4.25 / 96.33 ms │            85.05 / 87.85 ±3.22 / 93.45 ms │     no change │
│ QQuery 39 │        114.96 / 117.88 ±4.82 / 127.51 ms │         114.76 / 115.71 ±0.60 / 116.31 ms │     no change │
│ QQuery 40 │        101.41 / 107.28 ±3.90 / 112.27 ms │         101.64 / 108.54 ±5.25 / 117.51 ms │     no change │
│ QQuery 41 │           13.83 / 14.05 ±0.26 / 14.56 ms │            13.86 / 14.04 ±0.25 / 14.52 ms │     no change │
│ QQuery 42 │        106.47 / 108.15 ±3.10 / 114.34 ms │         105.96 / 106.56 ±0.42 / 107.19 ms │     no change │
│ QQuery 43 │              5.58 / 5.70 ±0.18 / 6.07 ms │               5.45 / 6.48 ±1.68 / 9.83 ms │  1.14x slower │
│ QQuery 44 │           11.37 / 11.61 ±0.18 / 11.93 ms │            11.36 / 11.54 ±0.13 / 11.70 ms │     no change │
│ QQuery 45 │           48.33 / 48.76 ±0.40 / 49.46 ms │            47.35 / 48.01 ±0.38 / 48.38 ms │     no change │
│ QQuery 46 │              8.28 / 8.45 ±0.19 / 8.80 ms │               8.20 / 8.32 ±0.16 / 8.63 ms │     no change │
│ QQuery 47 │        663.20 / 670.06 ±5.21 / 679.35 ms │         662.71 / 664.57 ±2.68 / 669.82 ms │     no change │
│ QQuery 48 │        271.64 / 274.90 ±3.06 / 280.56 ms │         272.03 / 275.05 ±2.04 / 277.69 ms │     no change │
│ QQuery 49 │        247.66 / 249.08 ±1.62 / 252.23 ms │         247.51 / 247.86 ±0.30 / 248.30 ms │     no change │
│ QQuery 50 │        203.21 / 206.91 ±3.14 / 211.87 ms │         200.05 / 205.43 ±6.93 / 218.55 ms │     no change │
│ QQuery 51 │        175.76 / 178.15 ±2.49 / 182.97 ms │         173.85 / 177.60 ±3.42 / 184.05 ms │     no change │
│ QQuery 52 │        106.21 / 106.67 ±0.50 / 107.60 ms │         106.91 / 107.03 ±0.15 / 107.32 ms │     no change │
│ QQuery 53 │        101.02 / 102.64 ±2.54 / 107.67 ms │         101.25 / 102.83 ±2.04 / 106.84 ms │     no change │
│ QQuery 54 │        142.13 / 144.29 ±2.08 / 148.02 ms │         142.74 / 143.24 ±0.33 / 143.55 ms │     no change │
│ QQuery 55 │        105.42 / 106.74 ±0.83 / 107.61 ms │         105.59 / 106.67 ±1.10 / 108.69 ms │     no change │
│ QQuery 56 │        139.17 / 140.02 ±1.07 / 142.08 ms │         138.35 / 139.22 ±0.78 / 140.59 ms │     no change │
│ QQuery 57 │        162.18 / 164.60 ±2.97 / 170.24 ms │         162.15 / 162.81 ±0.48 / 163.59 ms │     no change │
│ QQuery 58 │        306.68 / 306.99 ±0.43 / 307.84 ms │         306.27 / 307.44 ±1.07 / 309.05 ms │     no change │
│ QQuery 59 │        191.42 / 193.10 ±1.73 / 196.21 ms │         190.89 / 192.03 ±1.48 / 194.94 ms │     no change │
│ QQuery 60 │        140.11 / 141.03 ±0.77 / 142.19 ms │         140.01 / 140.76 ±1.22 / 143.18 ms │     no change │
│ QQuery 61 │           13.01 / 13.15 ±0.18 / 13.50 ms │            13.00 / 13.19 ±0.22 / 13.60 ms │     no change │
│ QQuery 62 │        857.24 / 864.90 ±7.25 / 877.63 ms │         850.69 / 858.97 ±7.91 / 872.69 ms │     no change │
│ QQuery 63 │        101.44 / 102.69 ±1.52 / 105.61 ms │         101.57 / 101.94 ±0.33 / 102.36 ms │     no change │
│ QQuery 64 │        648.38 / 653.84 ±4.37 / 661.61 ms │         648.76 / 652.18 ±5.38 / 662.70 ms │     no change │
│ QQuery 65 │        238.61 / 240.93 ±2.57 / 245.62 ms │         239.09 / 240.44 ±1.06 / 241.56 ms │     no change │
│ QQuery 66 │        208.84 / 218.54 ±8.96 / 233.68 ms │         211.08 / 219.31 ±8.31 / 234.02 ms │     no change │
│ QQuery 67 │        288.71 / 296.33 ±6.23 / 305.38 ms │         287.84 / 295.30 ±8.49 / 308.26 ms │     no change │
│ QQuery 68 │              8.52 / 8.67 ±0.20 / 9.06 ms │               8.21 / 8.43 ±0.27 / 8.96 ms │     no change │
│ QQuery 69 │           96.39 / 97.08 ±0.59 / 98.04 ms │            96.11 / 96.93 ±0.74 / 98.25 ms │     no change │
│ QQuery 70 │       308.84 / 323.41 ±10.51 / 338.47 ms │        308.12 / 328.03 ±13.70 / 345.74 ms │     no change │
│ QQuery 71 │        132.28 / 134.11 ±3.00 / 140.08 ms │         131.26 / 132.90 ±2.06 / 136.90 ms │     no change │
│ QQuery 72 │        572.48 / 585.33 ±6.71 / 591.71 ms │         569.75 / 583.85 ±7.62 / 590.61 ms │     no change │
│ QQuery 73 │              6.46 / 6.66 ±0.22 / 7.08 ms │               6.45 / 6.60 ±0.24 / 7.07 ms │     no change │
│ QQuery 74 │        521.88 / 526.35 ±3.90 / 533.08 ms │         519.00 / 524.80 ±3.81 / 529.65 ms │     no change │
│ QQuery 75 │        265.36 / 266.97 ±2.33 / 271.58 ms │         264.25 / 265.98 ±1.97 / 269.71 ms │     no change │
│ QQuery 76 │        128.61 / 130.19 ±1.12 / 132.09 ms │         127.82 / 129.47 ±2.12 / 133.61 ms │     no change │
│ QQuery 77 │        184.91 / 186.64 ±1.66 / 189.68 ms │         185.25 / 187.51 ±1.32 / 189.01 ms │     no change │
│ QQuery 78 │        324.97 / 327.72 ±2.41 / 332.19 ms │         323.42 / 326.61 ±2.56 / 330.65 ms │     no change │
│ QQuery 79 │        224.66 / 227.16 ±1.78 / 229.92 ms │         223.90 / 226.84 ±2.42 / 230.85 ms │     no change │
│ QQuery 80 │        316.31 / 319.52 ±2.69 / 323.21 ms │         317.90 / 319.41 ±1.72 / 322.38 ms │     no change │
│ QQuery 81 │           24.72 / 25.30 ±0.35 / 25.70 ms │            25.47 / 25.70 ±0.21 / 26.02 ms │     no change │
│ QQuery 82 │           38.75 / 38.93 ±0.13 / 39.12 ms │            38.23 / 38.58 ±0.30 / 39.13 ms │     no change │
│ QQuery 83 │           36.19 / 36.53 ±0.27 / 36.88 ms │            35.91 / 36.30 ±0.28 / 36.69 ms │     no change │
│ QQuery 84 │           45.67 / 46.13 ±0.38 / 46.56 ms │            45.50 / 45.80 ±0.18 / 46.04 ms │     no change │
│ QQuery 85 │        139.35 / 140.89 ±1.60 / 143.98 ms │         138.67 / 139.19 ±0.54 / 140.22 ms │     no change │
│ QQuery 86 │           36.80 / 37.27 ±0.31 / 37.70 ms │            36.30 / 37.00 ±0.84 / 38.57 ms │     no change │
│ QQuery 87 │              3.41 / 3.52 ±0.17 / 3.85 ms │               3.45 / 3.50 ±0.09 / 3.68 ms │     no change │
│ QQuery 88 │           97.85 / 98.30 ±0.46 / 99.02 ms │           97.91 / 99.78 ±2.30 / 103.90 ms │     no change │
│ QQuery 89 │        115.42 / 117.87 ±2.92 / 123.52 ms │         114.81 / 116.50 ±2.05 / 120.15 ms │     no change │
│ QQuery 90 │           21.63 / 22.14 ±0.36 / 22.70 ms │            21.74 / 22.05 ±0.30 / 22.57 ms │     no change │
│ QQuery 91 │           57.19 / 57.89 ±0.69 / 59.14 ms │            57.24 / 57.63 ±0.23 / 57.91 ms │     no change │
│ QQuery 92 │           55.32 / 56.31 ±1.11 / 58.44 ms │            54.86 / 55.43 ±0.41 / 56.13 ms │     no change │
│ QQuery 93 │        178.34 / 179.17 ±0.84 / 180.62 ms │         178.32 / 179.18 ±0.74 / 180.06 ms │     no change │
│ QQuery 94 │           59.87 / 60.11 ±0.32 / 60.71 ms │            59.72 / 60.41 ±0.61 / 61.25 ms │     no change │
│ QQuery 95 │        124.73 / 125.72 ±0.78 / 127.05 ms │         123.65 / 124.96 ±0.77 / 125.85 ms │     no change │
│ QQuery 96 │           66.36 / 68.48 ±1.32 / 70.39 ms │            66.95 / 68.30 ±1.25 / 70.35 ms │     no change │
│ QQuery 97 │        114.87 / 116.16 ±1.18 / 117.66 ms │         114.72 / 116.52 ±1.05 / 117.37 ms │     no change │
│ QQuery 98 │        147.50 / 149.96 ±1.47 / 151.48 ms │         147.38 / 149.05 ±1.39 / 151.53 ms │     no change │
│ QQuery 99 │ 10664.06 / 10733.90 ±54.47 / 10820.77 ms │ 10618.68 / 10780.69 ±100.29 / 10908.55 ms │     no change │
└───────────┴──────────────────────────────────────────┴───────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                 ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                 │ 30158.61ms │
│ Total Time (right_semi_default)   │ 30141.51ms │
│ Average Time (HEAD)               │   304.63ms │
│ Average Time (right_semi_default) │   304.46ms │
│ Queries Faster                    │          1 │
│ Queries Slower                    │          1 │
│ Queries with No Change            │         97 │
│ Queries with Failure              │          0 │
└───────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	155.0s
Peak memory	6.3 GiB
Avg memory	5.5 GiB
CPU user	251.0s
CPU sys	8.1s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	155.0s
Peak memory	6.4 GiB
Avg memory	5.6 GiB
CPU user	250.7s
CPU sys	8.1s
Peak spill	0 B

File an issue against this benchmark runner

Dandandan · 2026-04-23T20:20:13Z

Lets keep it for now as is

github-actions Bot added optimizer Optimizer rules sqllogictest SQL Logic Tests (.slt) labels Apr 23, 2026

Dandandan closed this Apr 23, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: plan IN/EXISTS subquery decorrelation as RightSemi by default#21809

feat: plan IN/EXISTS subquery decorrelation as RightSemi by default#21809
Dandandan wants to merge 1 commit intoapache:mainfrom
Dandandan:right_semi_default

Dandandan commented Apr 23, 2026 •

edited

Loading

Uh oh!

Dandandan commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

Dandandan commented Apr 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

Dandandan commented Apr 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Which issue does this PR close?

Rationale for this change

Are these changes tested?

Are there any user-facing changes?

Uh oh!

Dandandan commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

adriangbot commented Apr 23, 2026

Uh oh!

Dandandan commented Apr 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Dandandan commented Apr 23, 2026 •

edited

Loading