Coalesce adjacent ranges in ParquetPushDecoder I/O requests by Dandandan · Pull Request #21370 · apache/datafusion

Dandandan · 2026-04-04T20:00:32Z

Which issue does this PR close?

N/A

Rationale for this change

The ParquetPushDecoder may request adjacent byte ranges (e.g. 4-534 and 534-1064) that can be merged into a single range (4-1064), reducing the number of object store I/O operations.

What changes are included in this PR?

Added coalesce_ranges() function that merges adjacent/overlapping byte ranges in-place
Applied coalescing to the NeedsData branch of the PushDecoderStreamState transition loop in opener.rs
Updated test expectations to reflect the reduced number of range requests

Are these changes tested?

Yes, existing tests in object_store_access.rs verify the coalesced ranges.

Are there any user-facing changes?

No API changes. Users may see improved I/O performance when reading Parquet files due to fewer object store requests.

🤖 Generated with Claude Code

The ParquetPushDecoder may request adjacent byte ranges (e.g. 4-534 and 534-1064) that can be merged into a single range (4-1064), reducing the number of object store I/O operations. Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>

Dandandan · 2026-04-04T20:06:48Z

run benchmarks

adriangbot · 2026-04-04T20:09:19Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187671509-811-pznvl 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:09:22Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187671509-812-425hw 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:09:27Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187671509-810-gtphw 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

Dandandan · 2026-04-04T20:23:11Z

run benchmarks

env:
    DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
    DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true

adriangbot · 2026-04-04T20:25:09Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187694388-813-fbzdv 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:25:10Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187694388-815-lwd8g 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:25:10Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187694388-814-rfrwv 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:25:52Z

🤖 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 coalesce-push-decoder-ranges
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃             coalesce-push-decoder-ranges ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.58 / 7.01 ±0.84 / 8.69 ms │              6.53 / 7.01 ±0.80 / 8.61 ms │     no change │
│ QQuery 2  │        147.01 / 147.59 ±0.32 / 147.85 ms │        143.44 / 144.21 ±0.66 / 145.22 ms │     no change │
│ QQuery 3  │        114.06 / 115.06 ±0.65 / 115.82 ms │        113.58 / 113.91 ±0.27 / 114.33 ms │     no change │
│ QQuery 4  │    1345.43 / 1363.75 ±16.17 / 1386.73 ms │    1308.16 / 1343.22 ±20.08 / 1370.78 ms │     no change │
│ QQuery 5  │        174.09 / 176.09 ±1.42 / 178.08 ms │        174.14 / 175.16 ±0.81 / 176.58 ms │     no change │
│ QQuery 6  │       836.56 / 871.54 ±19.62 / 894.96 ms │       832.39 / 857.82 ±20.71 / 885.60 ms │     no change │
│ QQuery 7  │        352.91 / 355.45 ±1.67 / 357.37 ms │        351.05 / 353.98 ±2.01 / 357.33 ms │     no change │
│ QQuery 8  │        115.83 / 117.53 ±1.72 / 120.54 ms │        114.64 / 115.64 ±0.99 / 117.18 ms │     no change │
│ QQuery 9  │        102.19 / 107.81 ±6.59 / 120.21 ms │        101.14 / 105.02 ±2.07 / 106.80 ms │     no change │
│ QQuery 10 │        106.31 / 108.84 ±1.27 / 109.76 ms │        106.42 / 107.06 ±0.43 / 107.76 ms │     no change │
│ QQuery 11 │        927.40 / 942.23 ±9.45 / 955.42 ms │       917.02 / 938.22 ±15.80 / 961.89 ms │     no change │
│ QQuery 12 │           45.69 / 47.08 ±0.84 / 47.84 ms │           44.66 / 45.73 ±1.53 / 48.75 ms │     no change │
│ QQuery 13 │        397.59 / 400.71 ±3.18 / 404.69 ms │        391.35 / 395.67 ±2.47 / 398.67 ms │     no change │
│ QQuery 14 │     1019.82 / 1026.97 ±5.58 / 1034.62 ms │      998.98 / 1013.35 ±9.58 / 1026.54 ms │     no change │
│ QQuery 15 │           15.59 / 16.98 ±1.16 / 18.94 ms │           15.20 / 16.02 ±0.57 / 16.68 ms │ +1.06x faster │
│ QQuery 16 │              7.06 / 7.65 ±0.59 / 8.68 ms │              7.01 / 7.30 ±0.27 / 7.66 ms │     no change │
│ QQuery 17 │        243.02 / 244.58 ±1.53 / 247.29 ms │        236.72 / 238.26 ±1.12 / 240.20 ms │     no change │
│ QQuery 18 │        128.18 / 129.79 ±1.94 / 133.51 ms │        128.52 / 129.57 ±0.83 / 130.88 ms │     no change │
│ QQuery 19 │        155.92 / 158.27 ±1.62 / 160.40 ms │        153.61 / 155.03 ±0.90 / 156.04 ms │     no change │
│ QQuery 20 │           13.70 / 14.57 ±0.72 / 15.48 ms │           13.83 / 14.20 ±0.43 / 15.04 ms │     no change │
│ QQuery 21 │           19.76 / 20.73 ±0.76 / 21.64 ms │           19.74 / 20.20 ±0.33 / 20.62 ms │     no change │
│ QQuery 22 │        489.29 / 493.30 ±4.31 / 501.21 ms │        488.12 / 492.64 ±2.88 / 495.91 ms │     no change │
│ QQuery 23 │       885.42 / 903.27 ±12.26 / 917.14 ms │        886.54 / 898.13 ±6.36 / 905.72 ms │     no change │
│ QQuery 24 │        393.73 / 395.56 ±1.19 / 396.77 ms │        390.63 / 394.47 ±2.64 / 398.86 ms │     no change │
│ QQuery 25 │        350.34 / 353.02 ±2.00 / 356.41 ms │        347.51 / 351.99 ±3.67 / 357.33 ms │     no change │
│ QQuery 26 │           81.00 / 82.83 ±1.31 / 85.04 ms │           81.98 / 84.40 ±1.86 / 87.38 ms │     no change │
│ QQuery 27 │              6.89 / 7.70 ±0.74 / 8.75 ms │              6.78 / 7.19 ±0.42 / 7.98 ms │ +1.07x faster │
│ QQuery 28 │        148.69 / 150.10 ±0.91 / 150.91 ms │        147.33 / 148.59 ±0.84 / 149.61 ms │     no change │
│ QQuery 29 │        291.90 / 295.25 ±2.97 / 300.76 ms │        294.29 / 295.54 ±1.16 / 297.42 ms │     no change │
│ QQuery 30 │           42.53 / 45.28 ±2.14 / 47.30 ms │           43.26 / 44.28 ±1.03 / 46.26 ms │     no change │
│ QQuery 31 │        171.09 / 173.48 ±1.96 / 176.16 ms │        170.46 / 171.91 ±1.02 / 173.29 ms │     no change │
│ QQuery 32 │           57.09 / 57.57 ±0.46 / 58.18 ms │           55.91 / 56.93 ±0.76 / 57.81 ms │     no change │
│ QQuery 33 │        142.82 / 143.67 ±1.03 / 145.42 ms │        139.83 / 141.40 ±1.04 / 142.73 ms │     no change │
│ QQuery 34 │              6.97 / 7.24 ±0.26 / 7.73 ms │              6.98 / 7.37 ±0.45 / 8.19 ms │     no change │
│ QQuery 35 │        106.88 / 110.32 ±1.90 / 112.64 ms │        106.51 / 108.40 ±2.41 / 113.12 ms │     no change │
│ QQuery 36 │              6.59 / 6.77 ±0.20 / 7.13 ms │              6.43 / 6.74 ±0.25 / 7.05 ms │     no change │
│ QQuery 37 │             8.58 / 9.31 ±0.88 / 10.97 ms │              8.16 / 8.58 ±0.26 / 8.90 ms │ +1.09x faster │
│ QQuery 38 │           85.95 / 89.19 ±2.85 / 93.76 ms │           86.49 / 87.44 ±0.73 / 88.50 ms │     no change │
│ QQuery 39 │        126.85 / 130.07 ±2.32 / 132.30 ms │        126.26 / 127.87 ±1.30 / 129.96 ms │     no change │
│ QQuery 40 │        115.63 / 120.17 ±4.76 / 129.04 ms │        110.10 / 114.25 ±3.97 / 121.53 ms │     no change │
│ QQuery 41 │           15.09 / 15.93 ±0.66 / 16.96 ms │           14.59 / 15.25 ±0.59 / 16.27 ms │     no change │
│ QQuery 42 │        107.45 / 109.72 ±1.78 / 112.60 ms │        106.87 / 108.02 ±1.27 / 109.71 ms │     no change │
│ QQuery 43 │              5.95 / 6.12 ±0.14 / 6.37 ms │              5.96 / 6.52 ±1.10 / 8.71 ms │  1.07x slower │
│ QQuery 44 │           11.30 / 11.69 ±0.24 / 11.95 ms │           11.70 / 12.10 ±0.43 / 12.92 ms │     no change │
│ QQuery 45 │           52.51 / 53.61 ±0.90 / 55.21 ms │           51.87 / 53.37 ±0.84 / 54.40 ms │     no change │
│ QQuery 46 │              8.36 / 8.95 ±0.56 / 9.97 ms │              8.42 / 8.88 ±0.31 / 9.35 ms │     no change │
│ QQuery 47 │        713.43 / 722.11 ±8.51 / 736.67 ms │       724.81 / 740.61 ±13.15 / 758.47 ms │     no change │
│ QQuery 48 │        288.65 / 292.07 ±2.90 / 296.55 ms │        277.75 / 284.13 ±4.98 / 292.22 ms │     no change │
│ QQuery 49 │        255.84 / 259.14 ±2.07 / 261.96 ms │        253.60 / 255.74 ±2.14 / 259.73 ms │     no change │
│ QQuery 50 │        229.10 / 234.14 ±3.66 / 239.70 ms │        228.52 / 236.76 ±5.69 / 242.95 ms │     no change │
│ QQuery 51 │        179.87 / 186.07 ±3.75 / 191.42 ms │        182.26 / 185.08 ±2.56 / 189.21 ms │     no change │
│ QQuery 52 │        107.43 / 108.76 ±1.19 / 110.83 ms │        107.00 / 107.61 ±0.54 / 108.33 ms │     no change │
│ QQuery 53 │        102.58 / 103.96 ±1.18 / 105.83 ms │        102.35 / 102.55 ±0.20 / 102.85 ms │     no change │
│ QQuery 54 │        146.00 / 147.69 ±1.53 / 150.41 ms │        145.98 / 147.05 ±0.84 / 148.54 ms │     no change │
│ QQuery 55 │        107.21 / 109.92 ±1.88 / 112.45 ms │        106.21 / 107.14 ±0.77 / 108.18 ms │     no change │
│ QQuery 56 │        141.05 / 142.59 ±0.86 / 143.50 ms │        139.03 / 140.71 ±1.26 / 142.27 ms │     no change │
│ QQuery 57 │        173.34 / 176.19 ±1.66 / 178.18 ms │        171.88 / 173.79 ±1.30 / 175.96 ms │     no change │
│ QQuery 58 │        300.21 / 306.74 ±4.86 / 314.87 ms │        287.84 / 295.75 ±5.24 / 302.58 ms │     no change │
│ QQuery 59 │        199.44 / 202.20 ±1.79 / 204.75 ms │        196.15 / 199.16 ±2.47 / 201.96 ms │     no change │
│ QQuery 60 │        144.87 / 145.76 ±1.53 / 148.82 ms │        141.47 / 144.53 ±2.23 / 147.56 ms │     no change │
│ QQuery 61 │           13.06 / 13.32 ±0.32 / 13.95 ms │           13.18 / 13.41 ±0.29 / 13.97 ms │     no change │
│ QQuery 62 │       912.04 / 931.23 ±19.47 / 960.69 ms │      915.88 / 965.17 ±45.13 / 1030.53 ms │     no change │
│ QQuery 63 │        105.79 / 107.02 ±1.33 / 109.38 ms │        104.53 / 110.71 ±8.11 / 126.58 ms │     no change │
│ QQuery 64 │        696.83 / 702.33 ±6.07 / 713.65 ms │        691.10 / 697.42 ±5.88 / 708.28 ms │     no change │
│ QQuery 65 │        251.47 / 259.77 ±4.79 / 265.91 ms │        248.73 / 252.24 ±3.16 / 257.84 ms │     no change │
│ QQuery 66 │       246.11 / 263.67 ±12.85 / 277.58 ms │       243.85 / 260.05 ±18.38 / 293.61 ms │     no change │
│ QQuery 67 │        308.82 / 315.31 ±5.74 / 325.47 ms │        313.15 / 318.56 ±5.14 / 327.52 ms │     no change │
│ QQuery 68 │            8.99 / 10.02 ±0.72 / 10.86 ms │             8.28 / 9.85 ±0.91 / 10.77 ms │     no change │
│ QQuery 69 │        103.34 / 105.19 ±2.08 / 108.57 ms │        100.65 / 102.88 ±2.00 / 105.86 ms │     no change │
│ QQuery 70 │        339.71 / 348.65 ±6.61 / 358.93 ms │        326.94 / 342.76 ±8.29 / 349.63 ms │     no change │
│ QQuery 71 │        133.93 / 136.75 ±2.41 / 139.94 ms │        133.70 / 135.27 ±1.29 / 137.00 ms │     no change │
│ QQuery 72 │       711.13 / 721.66 ±12.27 / 737.83 ms │       699.33 / 720.69 ±12.98 / 734.16 ms │     no change │
│ QQuery 73 │              6.68 / 7.86 ±0.96 / 9.07 ms │             7.06 / 9.49 ±2.40 / 13.82 ms │  1.21x slower │
│ QQuery 74 │       582.63 / 592.80 ±10.02 / 609.52 ms │       561.33 / 572.56 ±10.34 / 588.78 ms │     no change │
│ QQuery 75 │        278.06 / 280.10 ±1.66 / 281.65 ms │        276.61 / 277.63 ±0.97 / 279.37 ms │     no change │
│ QQuery 76 │        132.02 / 133.61 ±0.89 / 134.45 ms │        131.10 / 132.49 ±1.31 / 134.08 ms │     no change │
│ QQuery 77 │        188.05 / 190.11 ±1.98 / 193.66 ms │        187.13 / 187.73 ±0.53 / 188.67 ms │     no change │
│ QQuery 78 │        355.00 / 360.19 ±4.08 / 367.12 ms │        344.25 / 350.60 ±5.51 / 357.26 ms │     no change │
│ QQuery 79 │        233.51 / 236.56 ±2.02 / 239.10 ms │        231.14 / 233.12 ±2.41 / 237.83 ms │     no change │
│ QQuery 80 │        328.17 / 331.14 ±1.83 / 333.91 ms │        329.08 / 330.83 ±2.91 / 336.63 ms │     no change │
│ QQuery 81 │           26.46 / 27.45 ±0.94 / 28.59 ms │           26.68 / 27.13 ±0.49 / 27.83 ms │     no change │
│ QQuery 82 │        198.21 / 200.99 ±2.98 / 205.53 ms │        193.25 / 195.01 ±1.52 / 197.76 ms │     no change │
│ QQuery 83 │           39.53 / 40.50 ±0.87 / 41.87 ms │           39.28 / 39.89 ±0.33 / 40.25 ms │     no change │
│ QQuery 84 │           47.59 / 48.52 ±0.74 / 49.49 ms │           47.57 / 48.51 ±0.55 / 49.14 ms │     no change │
│ QQuery 85 │        147.39 / 148.34 ±0.83 / 149.63 ms │        148.23 / 149.43 ±0.87 / 150.79 ms │     no change │
│ QQuery 86 │           38.68 / 39.49 ±0.66 / 40.21 ms │           38.65 / 39.84 ±0.77 / 40.69 ms │     no change │
│ QQuery 87 │           88.14 / 90.51 ±3.41 / 97.30 ms │           85.55 / 88.50 ±2.55 / 93.15 ms │     no change │
│ QQuery 88 │        100.85 / 101.94 ±1.40 / 104.60 ms │          98.55 / 99.67 ±1.74 / 103.12 ms │     no change │
│ QQuery 89 │        118.36 / 120.88 ±1.84 / 123.53 ms │        116.96 / 118.64 ±0.88 / 119.42 ms │     no change │
│ QQuery 90 │           24.08 / 24.95 ±1.35 / 27.63 ms │           23.20 / 23.58 ±0.42 / 24.39 ms │ +1.06x faster │
│ QQuery 91 │           64.55 / 65.40 ±0.60 / 66.10 ms │           62.12 / 64.16 ±1.58 / 66.68 ms │     no change │
│ QQuery 92 │           57.33 / 58.14 ±0.69 / 59.30 ms │           56.02 / 57.63 ±1.23 / 59.69 ms │     no change │
│ QQuery 93 │        194.39 / 195.08 ±0.50 / 195.65 ms │        189.50 / 192.31 ±1.62 / 194.47 ms │     no change │
│ QQuery 94 │           60.32 / 61.65 ±0.91 / 63.01 ms │           59.19 / 61.05 ±1.07 / 62.26 ms │     no change │
│ QQuery 95 │        136.17 / 137.06 ±0.58 / 137.72 ms │        133.27 / 134.06 ±0.78 / 135.45 ms │     no change │
│ QQuery 96 │           73.83 / 74.96 ±0.73 / 76.14 ms │           73.98 / 75.02 ±0.79 / 76.18 ms │     no change │
│ QQuery 97 │        129.70 / 133.06 ±2.94 / 136.77 ms │        130.36 / 132.29 ±1.59 / 134.52 ms │     no change │
│ QQuery 98 │        153.21 / 156.57 ±2.46 / 160.56 ms │        148.79 / 152.18 ±2.25 / 154.90 ms │     no change │
│ QQuery 99 │ 10816.80 / 10833.75 ±12.85 / 10848.43 ms │ 10806.73 / 10851.58 ±39.02 / 10899.76 ms │     no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                           ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                           │ 31904.22ms │
│ Total Time (coalesce-push-decoder-ranges)   │ 31743.36ms │
│ Average Time (HEAD)                         │   322.26ms │
│ Average Time (coalesce-push-decoder-ranges) │   320.64ms │
│ Queries Faster                              │          4 │
│ Queries Slower                              │          2 │
│ Queries with No Change                      │         93 │
│ Queries with Failure                        │          0 │
└─────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	159.8s
Peak memory	5.7 GiB
Avg memory	4.7 GiB
CPU user	261.0s
CPU sys	17.1s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	159.0s
Peak memory	6.2 GiB
Avg memory	5.2 GiB
CPU user	259.5s
CPU sys	16.7s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:26:52Z

🤖 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 coalesce-push-decoder-ranges
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃          coalesce-push-decoder-ranges ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.19 / 4.51 ±6.47 / 17.44 ms │          1.23 / 4.53 ±6.44 / 17.42 ms │     no change │
│ QQuery 1  │        14.25 / 14.56 ±0.19 / 14.82 ms │        14.34 / 14.64 ±0.19 / 14.89 ms │     no change │
│ QQuery 2  │        44.17 / 44.50 ±0.22 / 44.77 ms │        43.72 / 44.13 ±0.43 / 44.92 ms │     no change │
│ QQuery 3  │        40.69 / 43.32 ±2.15 / 46.37 ms │        42.65 / 44.91 ±1.25 / 46.14 ms │     no change │
│ QQuery 4  │     294.38 / 299.75 ±3.58 / 305.45 ms │     297.73 / 308.71 ±8.00 / 322.06 ms │     no change │
│ QQuery 5  │     352.88 / 356.11 ±3.34 / 362.42 ms │     348.61 / 355.32 ±6.14 / 366.71 ms │     no change │
│ QQuery 6  │           5.44 / 6.24 ±0.69 / 7.41 ms │           5.64 / 6.13 ±0.39 / 6.59 ms │     no change │
│ QQuery 7  │        17.11 / 18.54 ±1.32 / 20.24 ms │        16.98 / 17.25 ±0.44 / 18.13 ms │ +1.07x faster │
│ QQuery 8  │     424.89 / 432.26 ±7.25 / 444.43 ms │     426.74 / 434.75 ±6.17 / 441.62 ms │     no change │
│ QQuery 9  │     647.45 / 660.18 ±7.62 / 668.92 ms │     647.45 / 662.11 ±9.98 / 676.74 ms │     no change │
│ QQuery 10 │        91.70 / 94.79 ±1.81 / 96.93 ms │       93.94 / 96.13 ±2.68 / 101.21 ms │     no change │
│ QQuery 11 │     105.25 / 106.73 ±1.31 / 108.86 ms │     106.79 / 108.70 ±1.34 / 110.94 ms │     no change │
│ QQuery 12 │     347.17 / 353.87 ±4.95 / 361.72 ms │     349.34 / 355.77 ±6.76 / 368.32 ms │     no change │
│ QQuery 13 │     464.85 / 476.87 ±9.10 / 491.77 ms │    473.66 / 493.35 ±17.21 / 518.02 ms │     no change │
│ QQuery 14 │     358.25 / 364.07 ±3.34 / 368.26 ms │     360.82 / 364.33 ±2.72 / 368.93 ms │     no change │
│ QQuery 15 │    365.35 / 385.30 ±18.64 / 409.75 ms │    363.51 / 382.86 ±23.29 / 425.88 ms │     no change │
│ QQuery 16 │    722.96 / 761.58 ±36.28 / 828.41 ms │    719.98 / 760.41 ±32.05 / 797.99 ms │     no change │
│ QQuery 17 │    718.24 / 734.02 ±14.82 / 761.91 ms │     713.29 / 723.62 ±7.88 / 735.80 ms │     no change │
│ QQuery 18 │ 1422.33 / 1486.85 ±47.37 / 1554.29 ms │ 1460.09 / 1503.37 ±34.00 / 1562.55 ms │     no change │
│ QQuery 19 │        35.12 / 38.66 ±5.77 / 50.04 ms │      35.88 / 51.72 ±30.25 / 112.20 ms │  1.34x slower │
│ QQuery 20 │    709.95 / 736.26 ±26.83 / 781.02 ms │    709.28 / 734.75 ±28.93 / 790.46 ms │     no change │
│ QQuery 21 │     762.41 / 765.60 ±3.08 / 770.36 ms │     760.92 / 765.03 ±3.24 / 769.07 ms │     no change │
│ QQuery 22 │  1131.59 / 1140.20 ±8.58 / 1152.87 ms │  1130.56 / 1137.98 ±3.84 / 1141.49 ms │     no change │
│ QQuery 23 │ 3056.99 / 3093.07 ±29.36 / 3142.04 ms │ 3088.51 / 3107.17 ±13.93 / 3129.37 ms │     no change │
│ QQuery 24 │     102.83 / 104.13 ±1.55 / 107.07 ms │     103.91 / 105.17 ±1.25 / 106.82 ms │     no change │
│ QQuery 25 │     138.60 / 141.59 ±1.92 / 144.52 ms │     140.41 / 142.63 ±2.09 / 146.03 ms │     no change │
│ QQuery 26 │      99.74 / 101.34 ±1.74 / 104.70 ms │     101.08 / 102.90 ±1.32 / 104.36 ms │     no change │
│ QQuery 27 │     852.76 / 858.59 ±5.53 / 868.33 ms │     846.07 / 852.90 ±4.66 / 860.72 ms │     no change │
│ QQuery 28 │ 7709.08 / 7768.02 ±31.98 / 7794.78 ms │ 7707.93 / 7764.83 ±39.15 / 7812.27 ms │     no change │
│ QQuery 29 │        50.44 / 55.19 ±5.38 / 63.52 ms │        49.96 / 54.51 ±4.43 / 61.00 ms │     no change │
│ QQuery 30 │     359.59 / 368.90 ±7.50 / 381.96 ms │     355.56 / 366.42 ±9.21 / 379.65 ms │     no change │
│ QQuery 31 │     371.92 / 389.83 ±9.82 / 401.82 ms │    357.98 / 378.71 ±17.54 / 406.81 ms │     no change │
│ QQuery 32 │ 1209.58 / 1254.37 ±38.30 / 1322.59 ms │ 1024.84 / 1052.94 ±27.57 / 1099.14 ms │ +1.19x faster │
│ QQuery 33 │ 1474.98 / 1519.15 ±54.95 / 1621.25 ms │ 1455.92 / 1471.37 ±10.87 / 1486.96 ms │     no change │
│ QQuery 34 │ 1476.31 / 1499.62 ±18.22 / 1531.43 ms │  1466.03 / 1483.53 ±9.39 / 1494.09 ms │     no change │
│ QQuery 35 │     392.38 / 400.97 ±9.40 / 414.02 ms │     387.67 / 395.18 ±7.87 / 406.52 ms │     no change │
│ QQuery 36 │     122.15 / 125.33 ±3.03 / 130.40 ms │     118.68 / 124.42 ±3.03 / 127.70 ms │     no change │
│ QQuery 37 │        47.96 / 50.25 ±1.63 / 52.89 ms │        49.43 / 50.69 ±0.99 / 52.42 ms │     no change │
│ QQuery 38 │        74.67 / 77.23 ±1.68 / 79.37 ms │        74.83 / 77.17 ±1.62 / 79.52 ms │     no change │
│ QQuery 39 │     207.90 / 219.52 ±8.94 / 234.82 ms │     207.02 / 223.43 ±8.65 / 232.51 ms │     no change │
│ QQuery 40 │        24.75 / 27.16 ±2.01 / 29.42 ms │        22.95 / 27.17 ±2.13 / 28.85 ms │     no change │
│ QQuery 41 │        19.79 / 21.68 ±1.44 / 24.10 ms │        21.25 / 21.59 ±0.33 / 22.18 ms │     no change │
│ QQuery 42 │        19.99 / 20.49 ±0.27 / 20.78 ms │        19.73 / 20.33 ±0.32 / 20.66 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                           ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                           │ 27421.20ms │
│ Total Time (coalesce-push-decoder-ranges)   │ 27193.56ms │
│ Average Time (HEAD)                         │   637.70ms │
│ Average Time (coalesce-push-decoder-ranges) │   632.41ms │
│ Queries Faster                              │          2 │
│ Queries Slower                              │          1 │
│ Queries with No Change                      │         40 │
│ Queries with Failure                        │          0 │
└─────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	138.3s
Peak memory	37.6 GiB
Avg memory	27.5 GiB
CPU user	1293.1s
CPU sys	98.0s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	137.1s
Peak memory	39.6 GiB
Avg memory	30.2 GiB
CPU user	1293.1s
CPU sys	87.6s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:39:34Z

🤖 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 coalesce-push-decoder-ranges
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃          coalesce-push-decoder-ranges ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.21 / 4.55 ±6.50 / 17.55 ms │          1.25 / 4.56 ±6.49 / 17.53 ms │     no change │
│ QQuery 1  │        15.47 / 15.86 ±0.20 / 16.07 ms │        15.47 / 15.94 ±0.28 / 16.31 ms │     no change │
│ QQuery 2  │        44.65 / 44.92 ±0.25 / 45.38 ms │        44.03 / 44.23 ±0.13 / 44.41 ms │     no change │
│ QQuery 3  │        42.84 / 45.03 ±1.49 / 46.47 ms │        40.27 / 40.62 ±0.32 / 41.19 ms │ +1.11x faster │
│ QQuery 4  │    293.61 / 310.10 ±11.00 / 325.53 ms │     296.08 / 305.24 ±7.01 / 315.79 ms │     no change │
│ QQuery 5  │     358.66 / 361.10 ±2.29 / 364.97 ms │     354.15 / 357.38 ±2.78 / 361.39 ms │     no change │
│ QQuery 6  │           6.38 / 6.93 ±0.47 / 7.50 ms │           5.84 / 6.26 ±0.27 / 6.68 ms │ +1.11x faster │
│ QQuery 7  │        22.94 / 24.22 ±1.37 / 26.38 ms │        22.24 / 22.55 ±0.25 / 22.99 ms │ +1.07x faster │
│ QQuery 8  │     440.40 / 446.89 ±5.28 / 453.83 ms │     424.15 / 438.34 ±9.61 / 449.25 ms │     no change │
│ QQuery 9  │    666.43 / 680.21 ±11.78 / 698.81 ms │     662.42 / 675.95 ±7.72 / 685.71 ms │     no change │
│ QQuery 10 │     120.22 / 124.04 ±2.45 / 127.64 ms │     120.59 / 123.93 ±2.51 / 128.23 ms │     no change │
│ QQuery 11 │     135.66 / 137.10 ±1.08 / 138.47 ms │     134.71 / 135.92 ±1.37 / 138.55 ms │     no change │
│ QQuery 12 │     389.52 / 400.30 ±6.76 / 410.17 ms │    384.75 / 396.37 ±10.87 / 415.09 ms │     no change │
│ QQuery 13 │    510.29 / 520.89 ±12.08 / 543.99 ms │    511.42 / 528.60 ±13.54 / 546.61 ms │     no change │
│ QQuery 14 │     406.02 / 414.03 ±4.59 / 419.40 ms │     409.66 / 418.56 ±7.89 / 427.82 ms │     no change │
│ QQuery 15 │    393.16 / 407.19 ±11.75 / 428.69 ms │    384.04 / 406.86 ±25.63 / 455.32 ms │     no change │
│ QQuery 16 │    750.37 / 767.42 ±13.16 / 789.61 ms │     759.01 / 771.44 ±8.33 / 782.23 ms │     no change │
│ QQuery 17 │     731.60 / 741.27 ±7.58 / 752.71 ms │     744.01 / 750.37 ±4.06 / 756.29 ms │     no change │
│ QQuery 18 │ 1478.38 / 1510.27 ±19.28 / 1535.02 ms │ 1491.27 / 1513.84 ±32.11 / 1577.02 ms │     no change │
│ QQuery 19 │       36.53 / 45.81 ±15.79 / 77.32 ms │      35.86 / 57.32 ±39.77 / 136.85 ms │  1.25x slower │
│ QQuery 20 │    724.65 / 741.00 ±17.89 / 765.65 ms │    718.82 / 742.32 ±28.29 / 796.09 ms │     no change │
│ QQuery 21 │    721.70 / 747.48 ±14.58 / 765.29 ms │    721.69 / 734.63 ±11.68 / 754.69 ms │     no change │
│ QQuery 22 │ 1157.15 / 1171.77 ±11.56 / 1186.97 ms │  1178.25 / 1182.68 ±3.09 / 1187.02 ms │     no change │
│ QQuery 23 │    391.15 / 404.35 ±10.84 / 419.45 ms │     396.39 / 404.32 ±6.60 / 416.15 ms │     no change │
│ QQuery 24 │     109.99 / 111.75 ±2.60 / 116.88 ms │     112.39 / 114.94 ±1.47 / 116.42 ms │     no change │
│ QQuery 25 │     175.79 / 180.56 ±4.00 / 186.59 ms │     177.50 / 183.99 ±4.63 / 191.27 ms │     no change │
│ QQuery 26 │     137.48 / 145.24 ±4.92 / 152.98 ms │     134.34 / 142.99 ±4.59 / 147.12 ms │     no change │
│ QQuery 27 │   974.68 / 985.86 ±17.35 / 1020.33 ms │   971.38 / 990.37 ±11.99 / 1007.64 ms │     no change │
│ QQuery 28 │ 7872.21 / 7964.10 ±47.19 / 8003.69 ms │ 7891.84 / 7957.81 ±34.54 / 7990.91 ms │     no change │
│ QQuery 29 │        50.68 / 54.75 ±3.60 / 60.31 ms │        50.68 / 56.81 ±6.21 / 66.88 ms │     no change │
│ QQuery 30 │     373.01 / 380.33 ±4.50 / 386.62 ms │     382.91 / 393.46 ±7.16 / 402.15 ms │     no change │
│ QQuery 31 │     365.29 / 372.08 ±5.96 / 382.34 ms │    370.57 / 389.47 ±14.03 / 412.88 ms │     no change │
│ QQuery 32 │ 1276.44 / 1360.22 ±49.50 / 1412.50 ms │ 1077.47 / 1136.04 ±63.16 / 1251.23 ms │ +1.20x faster │
│ QQuery 33 │ 1646.38 / 1679.88 ±34.40 / 1729.09 ms │ 1510.83 / 1534.99 ±18.63 / 1559.57 ms │ +1.09x faster │
│ QQuery 34 │ 1556.12 / 1641.23 ±62.84 / 1748.58 ms │ 1494.67 / 1545.73 ±37.84 / 1607.18 ms │ +1.06x faster │
│ QQuery 35 │    421.93 / 447.13 ±26.43 / 492.88 ms │    408.81 / 420.37 ±16.21 / 452.18 ms │ +1.06x faster │
│ QQuery 36 │     108.51 / 111.15 ±2.32 / 114.88 ms │      98.96 / 109.16 ±7.89 / 123.22 ms │     no change │
│ QQuery 37 │        53.17 / 57.75 ±2.91 / 60.86 ms │        54.30 / 58.53 ±3.34 / 63.50 ms │     no change │
│ QQuery 38 │        54.31 / 60.50 ±3.45 / 63.76 ms │        56.76 / 59.28 ±2.34 / 63.37 ms │     no change │
│ QQuery 39 │     181.34 / 189.00 ±4.24 / 192.90 ms │     173.21 / 183.94 ±9.51 / 201.43 ms │     no change │
│ QQuery 40 │        30.82 / 32.09 ±1.44 / 33.98 ms │        32.79 / 34.15 ±1.18 / 35.92 ms │  1.06x slower │
│ QQuery 41 │        25.65 / 28.61 ±2.16 / 32.32 ms │        24.90 / 26.96 ±1.31 / 28.68 ms │ +1.06x faster │
│ QQuery 42 │        22.08 / 22.82 ±0.42 / 23.33 ms │        22.38 / 22.86 ±0.53 / 23.80 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                           ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                           │ 25897.76ms │
│ Total Time (coalesce-push-decoder-ranges)   │ 25440.09ms │
│ Average Time (HEAD)                         │   602.27ms │
│ Average Time (coalesce-push-decoder-ranges) │   591.63ms │
│ Queries Faster                              │          8 │
│ Queries Slower                              │          2 │
│ Queries with No Change                      │         33 │
│ Queries with Failure                        │          0 │
└─────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	130.3s
Peak memory	39.3 GiB
Avg memory	28.3 GiB
CPU user	1170.0s
CPU sys	118.4s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	128.2s
Peak memory	37.9 GiB
Avg memory	28.1 GiB
CPU user	1167.1s
CPU sys	100.8s
Peak spill	0 B

File an issue against this benchmark runner

Dandandan · 2026-04-04T20:43:12Z

run benchmarks

env:
    DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
    DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true

adriangbot · 2026-04-04T20:43:55Z

🤖 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 coalesce-push-decoder-ranges
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                      HEAD ┃              coalesce-push-decoder-ranges ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │               6.25 / 6.73 ±0.85 / 8.42 ms │               6.20 / 6.67 ±0.81 / 8.29 ms │     no change │
│ QQuery 2  │        111.63 / 122.41 ±19.34 / 161.01 ms │        111.65 / 131.43 ±23.58 / 162.01 ms │  1.07x slower │
│ QQuery 3  │         108.64 / 109.99 ±1.23 / 111.96 ms │         108.50 / 109.57 ±0.81 / 110.41 ms │     no change │
│ QQuery 4  │     1066.62 / 1093.18 ±15.39 / 1109.35 ms │      1091.77 / 1101.43 ±8.51 / 1115.49 ms │     no change │
│ QQuery 5  │         195.08 / 197.24 ±1.31 / 199.10 ms │         192.33 / 196.15 ±1.96 / 197.59 ms │     no change │
│ QQuery 6  │         248.13 / 257.32 ±8.34 / 270.01 ms │         254.90 / 263.67 ±4.70 / 268.74 ms │     no change │
│ QQuery 7  │         327.72 / 336.64 ±4.64 / 341.32 ms │         321.01 / 326.63 ±6.17 / 338.26 ms │     no change │
│ QQuery 8  │         160.31 / 165.80 ±3.27 / 169.20 ms │         159.75 / 161.79 ±1.84 / 165.13 ms │     no change │
│ QQuery 9  │        222.22 / 238.04 ±14.86 / 265.53 ms │         243.12 / 247.88 ±3.74 / 254.49 ms │     no change │
│ QQuery 10 │         171.04 / 177.98 ±4.17 / 183.49 ms │         167.30 / 175.14 ±4.97 / 181.63 ms │     no change │
│ QQuery 11 │        694.10 / 709.99 ±11.55 / 722.78 ms │         708.09 / 712.63 ±4.62 / 721.29 ms │     no change │
│ QQuery 12 │            38.09 / 39.96 ±1.79 / 43.14 ms │            38.71 / 40.23 ±1.34 / 42.29 ms │     no change │
│ QQuery 13 │        536.50 / 556.30 ±11.42 / 568.31 ms │         552.14 / 559.05 ±3.66 / 562.14 ms │     no change │
│ QQuery 14 │         891.29 / 900.01 ±9.69 / 918.89 ms │         898.20 / 906.90 ±7.51 / 916.08 ms │     no change │
│ QQuery 15 │            18.75 / 20.99 ±2.44 / 25.24 ms │            19.26 / 20.80 ±0.80 / 21.54 ms │     no change │
│ QQuery 16 │               7.02 / 7.24 ±0.18 / 7.45 ms │               7.03 / 7.24 ±0.21 / 7.57 ms │     no change │
│ QQuery 17 │         196.45 / 206.44 ±5.18 / 211.35 ms │         199.77 / 202.10 ±1.82 / 205.31 ms │     no change │
│ QQuery 18 │        415.44 / 468.06 ±27.88 / 492.55 ms │        399.41 / 444.91 ±33.52 / 487.96 ms │     no change │
│ QQuery 19 │         140.54 / 144.85 ±2.51 / 147.54 ms │         138.04 / 141.45 ±1.93 / 143.28 ms │     no change │
│ QQuery 20 │            15.59 / 17.29 ±1.20 / 19.05 ms │            15.28 / 16.04 ±0.44 / 16.59 ms │ +1.08x faster │
│ QQuery 21 │            26.53 / 27.73 ±1.44 / 30.34 ms │            27.40 / 28.27 ±0.87 / 29.88 ms │     no change │
│ QQuery 22 │         507.98 / 511.65 ±2.77 / 514.73 ms │         512.94 / 521.53 ±8.36 / 537.27 ms │     no change │
│ QQuery 23 │     1232.72 / 1278.87 ±68.23 / 1413.44 ms │     1227.82 / 1251.52 ±15.37 / 1275.18 ms │     no change │
│ QQuery 24 │         123.57 / 126.17 ±2.05 / 129.61 ms │         121.37 / 124.82 ±2.56 / 128.63 ms │     no change │
│ QQuery 25 │         298.40 / 302.55 ±2.83 / 306.06 ms │         304.91 / 307.64 ±3.49 / 314.17 ms │     no change │
│ QQuery 26 │         143.02 / 150.55 ±7.00 / 162.18 ms │         146.89 / 149.72 ±2.02 / 152.01 ms │     no change │
│ QQuery 27 │               6.54 / 7.26 ±0.77 / 8.57 ms │               6.78 / 7.06 ±0.30 / 7.61 ms │     no change │
│ QQuery 28 │         219.85 / 224.70 ±3.82 / 230.64 ms │         220.26 / 225.08 ±3.23 / 230.06 ms │     no change │
│ QQuery 29 │         251.65 / 255.40 ±2.04 / 257.61 ms │         246.63 / 250.34 ±3.18 / 256.01 ms │     no change │
│ QQuery 30 │            57.10 / 60.49 ±3.48 / 65.35 ms │            56.73 / 59.80 ±1.93 / 62.13 ms │     no change │
│ QQuery 31 │         177.09 / 178.82 ±1.59 / 181.44 ms │         177.32 / 178.78 ±1.94 / 182.52 ms │     no change │
│ QQuery 32 │            52.87 / 54.38 ±0.97 / 55.67 ms │            52.14 / 53.07 ±0.66 / 53.89 ms │     no change │
│ QQuery 33 │         132.54 / 134.51 ±2.61 / 139.65 ms │         127.78 / 130.28 ±1.55 / 131.89 ms │     no change │
│ QQuery 34 │               7.27 / 7.56 ±0.23 / 7.91 ms │               7.01 / 7.25 ±0.19 / 7.47 ms │     no change │
│ QQuery 35 │         147.51 / 150.37 ±1.99 / 152.35 ms │         142.44 / 149.02 ±3.47 / 152.57 ms │     no change │
│ QQuery 36 │               6.45 / 7.09 ±1.05 / 9.17 ms │               6.56 / 6.72 ±0.20 / 7.12 ms │ +1.05x faster │
│ QQuery 37 │               4.23 / 4.37 ±0.11 / 4.54 ms │               4.34 / 4.85 ±0.82 / 6.47 ms │  1.11x slower │
│ QQuery 38 │         107.79 / 110.91 ±3.73 / 118.13 ms │         103.78 / 105.90 ±1.57 / 107.48 ms │     no change │
│ QQuery 39 │         136.26 / 138.89 ±2.37 / 143.10 ms │         136.54 / 139.32 ±3.43 / 145.99 ms │     no change │
│ QQuery 40 │         133.70 / 141.95 ±8.47 / 157.10 ms │         136.52 / 140.83 ±5.48 / 151.20 ms │     no change │
│ QQuery 41 │            14.15 / 15.35 ±0.76 / 16.50 ms │            14.25 / 16.36 ±1.55 / 18.44 ms │  1.07x slower │
│ QQuery 42 │         112.96 / 115.42 ±1.71 / 117.63 ms │         106.63 / 112.26 ±3.77 / 115.77 ms │     no change │
│ QQuery 43 │               5.82 / 6.00 ±0.22 / 6.37 ms │               5.88 / 6.66 ±0.90 / 8.24 ms │  1.11x slower │
│ QQuery 44 │            11.04 / 12.39 ±1.57 / 15.24 ms │            10.95 / 11.99 ±0.87 / 13.34 ms │     no change │
│ QQuery 45 │            44.69 / 46.09 ±1.35 / 48.56 ms │            43.64 / 46.36 ±1.42 / 47.51 ms │     no change │
│ QQuery 46 │               8.23 / 8.66 ±0.38 / 9.19 ms │               8.33 / 8.59 ±0.26 / 9.07 ms │     no change │
│ QQuery 47 │        770.49 / 793.29 ±14.05 / 811.16 ms │         776.35 / 787.08 ±8.01 / 800.31 ms │     no change │
│ QQuery 48 │         486.76 / 500.20 ±7.81 / 511.07 ms │         479.25 / 487.52 ±6.44 / 495.58 ms │     no change │
│ QQuery 49 │         280.00 / 285.92 ±6.40 / 298.08 ms │         270.46 / 279.69 ±5.75 / 288.35 ms │     no change │
│ QQuery 50 │        529.85 / 548.68 ±10.74 / 560.24 ms │         527.34 / 538.73 ±6.86 / 548.75 ms │     no change │
│ QQuery 51 │         223.95 / 226.60 ±3.01 / 230.62 ms │         216.89 / 220.73 ±2.75 / 225.35 ms │     no change │
│ QQuery 52 │         114.36 / 116.20 ±1.41 / 117.90 ms │         107.46 / 113.20 ±4.55 / 117.32 ms │     no change │
│ QQuery 53 │         142.66 / 146.31 ±2.96 / 150.18 ms │         137.21 / 142.34 ±3.15 / 145.29 ms │     no change │
│ QQuery 54 │         127.49 / 129.79 ±2.42 / 133.02 ms │         118.78 / 125.05 ±4.87 / 133.74 ms │     no change │
│ QQuery 55 │         107.90 / 114.67 ±3.47 / 117.47 ms │         102.14 / 110.07 ±6.02 / 116.60 ms │     no change │
│ QQuery 56 │         133.40 / 135.19 ±1.36 / 137.49 ms │         127.20 / 129.93 ±1.61 / 132.09 ms │     no change │
│ QQuery 57 │         188.72 / 193.09 ±2.32 / 195.58 ms │         191.99 / 192.98 ±0.89 / 194.51 ms │     no change │
│ QQuery 58 │         373.73 / 381.42 ±7.10 / 392.09 ms │         363.72 / 369.84 ±5.96 / 379.14 ms │     no change │
│ QQuery 59 │         275.91 / 279.61 ±2.71 / 283.58 ms │         267.47 / 274.66 ±5.37 / 279.36 ms │     no change │
│ QQuery 60 │         138.33 / 140.95 ±2.10 / 144.65 ms │         132.94 / 135.29 ±1.30 / 136.58 ms │     no change │
│ QQuery 61 │            11.98 / 12.39 ±0.36 / 12.90 ms │            11.99 / 12.57 ±0.40 / 13.13 ms │     no change │
│ QQuery 62 │       890.97 / 965.68 ±60.39 / 1064.63 ms │       921.99 / 976.75 ±35.42 / 1032.52 ms │     no change │
│ QQuery 63 │         136.08 / 145.78 ±6.99 / 153.50 ms │         138.43 / 145.58 ±7.24 / 158.50 ms │     no change │
│ QQuery 64 │ 28186.14 / 28810.11 ±633.31 / 29818.16 ms │ 28180.33 / 28653.16 ±342.93 / 29085.33 ms │     no change │
│ QQuery 65 │         359.59 / 368.77 ±6.02 / 377.60 ms │        354.41 / 367.83 ±10.28 / 384.45 ms │     no change │
│ QQuery 66 │         194.67 / 201.42 ±4.33 / 207.67 ms │         194.26 / 196.54 ±1.68 / 199.33 ms │     no change │
│ QQuery 67 │        495.61 / 514.78 ±11.50 / 529.63 ms │         521.82 / 530.96 ±7.76 / 539.51 ms │     no change │
│ QQuery 68 │             8.86 / 10.17 ±1.64 / 13.40 ms │              8.56 / 9.40 ±0.73 / 10.43 ms │ +1.08x faster │
│ QQuery 69 │         174.12 / 180.99 ±7.74 / 195.35 ms │         162.47 / 177.03 ±7.76 / 183.62 ms │     no change │
│ QQuery 70 │         422.71 / 434.04 ±6.45 / 442.49 ms │        424.26 / 438.65 ±11.59 / 455.14 ms │     no change │
│ QQuery 71 │         128.50 / 131.42 ±1.82 / 133.69 ms │         124.84 / 127.88 ±3.59 / 133.62 ms │     no change │
│ QQuery 72 │     1213.28 / 1233.21 ±17.37 / 1262.54 ms │     1203.93 / 1213.30 ±10.37 / 1232.92 ms │     no change │
│ QQuery 73 │               7.23 / 8.37 ±0.82 / 9.51 ms │               6.54 / 7.11 ±0.87 / 8.81 ms │ +1.18x faster │
│ QQuery 74 │         510.71 / 512.10 ±1.57 / 515.05 ms │         494.60 / 509.76 ±8.00 / 518.28 ms │     no change │
│ QQuery 75 │         283.52 / 287.02 ±2.16 / 289.76 ms │         278.92 / 283.50 ±4.56 / 291.10 ms │     no change │
│ QQuery 76 │         277.55 / 283.09 ±3.43 / 287.20 ms │         279.59 / 283.41 ±2.93 / 287.92 ms │     no change │
│ QQuery 77 │         223.70 / 228.16 ±3.48 / 233.86 ms │         220.99 / 225.62 ±4.69 / 234.16 ms │     no change │
│ QQuery 78 │         302.95 / 309.78 ±4.22 / 315.61 ms │         304.96 / 311.61 ±6.21 / 321.48 ms │     no change │
│ QQuery 79 │         260.47 / 263.61 ±2.68 / 267.20 ms │         252.18 / 260.22 ±5.12 / 267.83 ms │     no change │
│ QQuery 80 │         315.55 / 326.65 ±5.58 / 330.46 ms │         313.59 / 319.46 ±3.79 / 323.32 ms │     no change │
│ QQuery 81 │            32.98 / 34.64 ±1.09 / 35.66 ms │            31.64 / 32.96 ±1.81 / 36.55 ms │     no change │
│ QQuery 82 │         194.75 / 197.41 ±2.14 / 200.96 ms │         190.85 / 194.48 ±2.70 / 198.22 ms │     no change │
│ QQuery 83 │            45.60 / 46.74 ±0.95 / 48.05 ms │            45.70 / 46.94 ±0.79 / 48.00 ms │     no change │
│ QQuery 84 │            64.08 / 65.39 ±1.16 / 67.25 ms │            63.60 / 65.43 ±1.17 / 66.63 ms │     no change │
│ QQuery 85 │         281.07 / 291.79 ±6.10 / 298.42 ms │         270.54 / 286.13 ±9.87 / 301.66 ms │     no change │
│ QQuery 86 │            47.12 / 50.01 ±1.78 / 52.20 ms │            45.74 / 48.76 ±1.69 / 50.91 ms │     no change │
│ QQuery 87 │          99.82 / 107.57 ±4.89 / 111.71 ms │          99.62 / 105.28 ±5.91 / 114.31 ms │     no change │
│ QQuery 88 │         121.01 / 130.82 ±6.14 / 139.45 ms │         132.59 / 133.52 ±0.83 / 135.03 ms │     no change │
│ QQuery 89 │         152.86 / 160.17 ±4.74 / 166.25 ms │         155.53 / 161.34 ±4.66 / 167.39 ms │     no change │
│ QQuery 90 │            26.96 / 28.08 ±0.87 / 29.62 ms │            23.73 / 27.24 ±2.06 / 29.67 ms │     no change │
│ QQuery 91 │          96.22 / 104.31 ±4.55 / 110.11 ms │          99.74 / 105.86 ±3.19 / 109.00 ms │     no change │
│ QQuery 92 │            53.60 / 54.11 ±0.58 / 55.22 ms │            53.02 / 54.13 ±1.14 / 55.73 ms │     no change │
│ QQuery 93 │         192.19 / 194.82 ±2.36 / 199.17 ms │         190.27 / 192.23 ±1.84 / 194.46 ms │     no change │
│ QQuery 94 │            64.05 / 64.94 ±0.78 / 66.12 ms │            59.57 / 62.47 ±2.06 / 64.49 ms │     no change │
│ QQuery 95 │         143.90 / 146.25 ±1.18 / 146.92 ms │         137.90 / 145.42 ±8.05 / 161.01 ms │     no change │
│ QQuery 96 │            75.46 / 86.49 ±5.63 / 91.28 ms │            76.90 / 86.13 ±4.63 / 88.99 ms │     no change │
│ QQuery 97 │         160.34 / 162.19 ±1.62 / 164.75 ms │         154.97 / 159.69 ±3.80 / 165.69 ms │     no change │
│ QQuery 98 │         121.30 / 123.92 ±1.99 / 126.18 ms │         120.47 / 122.82 ±1.37 / 124.52 ms │     no change │
│ QQuery 99 │  10778.54 / 10810.44 ±29.69 / 10861.56 ms │  10802.91 / 10848.90 ±25.78 / 10880.55 ms │     no change │
└───────────┴───────────────────────────────────────────┴───────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                           ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                           │ 61934.14ms │
│ Total Time (coalesce-push-decoder-ranges)   │ 61652.88ms │
│ Average Time (HEAD)                         │   625.60ms │
│ Average Time (coalesce-push-decoder-ranges) │   622.76ms │
│ Queries Faster                              │          4 │
│ Queries Slower                              │          4 │
│ Queries with No Change                      │         91 │
│ Queries with Failure                        │          0 │
└─────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	310.1s
Peak memory	39.0 GiB
Avg memory	27.7 GiB
CPU user	611.8s
CPU sys	27.2s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	308.7s
Peak memory	36.7 GiB
Avg memory	25.8 GiB
CPU user	607.1s
CPU sys	26.2s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:44:17Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187719183-817-vvxv6 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:45:52Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187719183-818-kdglh 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:45:53Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4187719183-816-mr2ww 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 coalesce-push-decoder-ranges (9f06337) to c17c87c (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-04T20:59:47Z

🤖 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 coalesce-push-decoder-ranges
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃          coalesce-push-decoder-ranges ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.21 / 4.53 ±6.50 / 17.53 ms │          1.21 / 4.50 ±6.44 / 17.39 ms │     no change │
│ QQuery 1  │        15.62 / 16.00 ±0.29 / 16.50 ms │        15.46 / 15.78 ±0.21 / 16.12 ms │     no change │
│ QQuery 2  │        43.70 / 45.18 ±1.49 / 48.04 ms │        44.28 / 44.65 ±0.24 / 44.99 ms │     no change │
│ QQuery 3  │        41.37 / 45.05 ±2.61 / 48.23 ms │        43.77 / 45.46 ±1.35 / 47.03 ms │     no change │
│ QQuery 4  │     293.54 / 301.80 ±6.80 / 313.24 ms │     284.01 / 297.73 ±8.87 / 307.12 ms │     no change │
│ QQuery 5  │     349.84 / 352.66 ±2.43 / 355.68 ms │     347.75 / 352.07 ±3.82 / 358.24 ms │     no change │
│ QQuery 6  │           5.82 / 6.13 ±0.32 / 6.72 ms │           5.92 / 6.21 ±0.32 / 6.83 ms │     no change │
│ QQuery 7  │        21.79 / 22.16 ±0.26 / 22.53 ms │        21.57 / 21.75 ±0.11 / 21.91 ms │     no change │
│ QQuery 8  │     424.44 / 432.41 ±6.71 / 444.46 ms │    431.85 / 452.91 ±14.84 / 472.33 ms │     no change │
│ QQuery 9  │     636.09 / 641.94 ±5.01 / 650.09 ms │     632.71 / 645.25 ±7.35 / 654.97 ms │     no change │
│ QQuery 10 │     117.96 / 120.19 ±1.90 / 122.42 ms │     118.66 / 121.33 ±2.90 / 126.95 ms │     no change │
│ QQuery 11 │     129.71 / 131.51 ±1.26 / 133.04 ms │     131.48 / 133.36 ±1.69 / 136.39 ms │     no change │
│ QQuery 12 │     382.13 / 387.68 ±3.06 / 390.82 ms │    377.40 / 398.52 ±25.62 / 443.64 ms │     no change │
│ QQuery 13 │    476.07 / 497.00 ±10.64 / 503.89 ms │    483.53 / 508.70 ±15.71 / 525.21 ms │     no change │
│ QQuery 14 │     392.35 / 402.55 ±5.35 / 406.70 ms │     389.60 / 397.17 ±5.37 / 404.11 ms │     no change │
│ QQuery 15 │    365.40 / 385.39 ±25.58 / 435.28 ms │     359.80 / 367.59 ±5.68 / 377.42 ms │     no change │
│ QQuery 16 │    734.84 / 753.30 ±16.95 / 777.09 ms │    728.83 / 736.84 ±10.29 / 756.71 ms │     no change │
│ QQuery 17 │     714.98 / 723.72 ±7.10 / 736.20 ms │     718.30 / 723.78 ±5.12 / 733.52 ms │     no change │
│ QQuery 18 │ 1432.91 / 1503.08 ±61.99 / 1609.05 ms │ 1421.73 / 1476.23 ±32.82 / 1513.40 ms │     no change │
│ QQuery 19 │        38.07 / 40.09 ±2.64 / 45.30 ms │       36.34 / 44.61 ±14.62 / 73.81 ms │  1.11x slower │
│ QQuery 20 │    712.66 / 732.34 ±21.45 / 768.77 ms │    716.65 / 728.79 ±13.36 / 753.50 ms │     no change │
│ QQuery 21 │    720.56 / 742.63 ±17.47 / 768.83 ms │    754.08 / 769.77 ±18.05 / 804.05 ms │     no change │
│ QQuery 22 │ 1153.76 / 1172.05 ±16.32 / 1201.03 ms │ 1152.55 / 1174.01 ±17.10 / 1200.41 ms │     no change │
│ QQuery 23 │     395.16 / 403.57 ±9.00 / 417.54 ms │    383.28 / 398.30 ±13.18 / 422.59 ms │     no change │
│ QQuery 24 │     107.57 / 109.32 ±0.95 / 110.21 ms │     106.58 / 109.53 ±2.24 / 112.18 ms │     no change │
│ QQuery 25 │     172.74 / 177.29 ±3.70 / 182.17 ms │     176.58 / 180.83 ±3.68 / 186.64 ms │     no change │
│ QQuery 26 │     136.52 / 141.86 ±3.43 / 146.86 ms │     137.22 / 140.82 ±1.87 / 142.58 ms │     no change │
│ QQuery 27 │   960.43 / 983.25 ±19.62 / 1017.45 ms │   953.01 / 991.26 ±25.22 / 1023.33 ms │     no change │
│ QQuery 28 │ 7846.68 / 7925.43 ±44.89 / 7981.01 ms │ 7870.18 / 7914.50 ±23.21 / 7932.83 ms │     no change │
│ QQuery 29 │        50.33 / 56.87 ±6.78 / 68.99 ms │        50.45 / 53.71 ±3.67 / 59.80 ms │ +1.06x faster │
│ QQuery 30 │     370.40 / 380.67 ±9.24 / 396.12 ms │     371.76 / 376.78 ±6.00 / 388.52 ms │     no change │
│ QQuery 31 │     363.45 / 369.66 ±4.82 / 375.29 ms │     362.24 / 368.37 ±4.43 / 373.67 ms │     no change │
│ QQuery 32 │ 1045.09 / 1067.09 ±30.01 / 1122.61 ms │ 1052.47 / 1095.71 ±48.04 / 1175.36 ms │     no change │
│ QQuery 33 │ 1458.67 / 1476.36 ±10.55 / 1487.92 ms │ 1461.34 / 1491.87 ±19.29 / 1515.95 ms │     no change │
│ QQuery 34 │ 1471.29 / 1498.56 ±22.44 / 1522.07 ms │  1484.04 / 1489.65 ±5.15 / 1496.30 ms │     no change │
│ QQuery 35 │     378.07 / 388.53 ±7.55 / 399.22 ms │     384.82 / 391.75 ±7.20 / 404.19 ms │     no change │
│ QQuery 36 │     106.57 / 109.21 ±2.21 / 112.95 ms │     106.44 / 111.07 ±3.55 / 116.98 ms │     no change │
│ QQuery 37 │        51.71 / 56.86 ±3.90 / 61.84 ms │        52.03 / 55.07 ±2.08 / 57.40 ms │     no change │
│ QQuery 38 │        52.54 / 54.40 ±1.37 / 56.22 ms │        56.13 / 58.27 ±2.10 / 61.92 ms │  1.07x slower │
│ QQuery 39 │     179.07 / 184.75 ±4.23 / 190.37 ms │     183.51 / 189.32 ±5.10 / 197.57 ms │     no change │
│ QQuery 40 │        29.45 / 33.08 ±4.05 / 40.70 ms │        30.31 / 32.36 ±2.16 / 35.01 ms │     no change │
│ QQuery 41 │        25.39 / 27.25 ±1.01 / 28.38 ms │        27.30 / 28.13 ±1.49 / 31.10 ms │     no change │
│ QQuery 42 │        21.13 / 22.09 ±0.48 / 22.42 ms │        22.51 / 22.90 ±0.28 / 23.30 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                           ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                           │ 24925.49ms │
│ Total Time (coalesce-push-decoder-ranges)   │ 24967.23ms │
│ Average Time (HEAD)                         │   579.66ms │
│ Average Time (coalesce-push-decoder-ranges) │   580.63ms │
│ Queries Faster                              │          1 │
│ Queries Slower                              │          2 │
│ Queries with No Change                      │         40 │
│ Queries with Failure                        │          0 │
└─────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	125.8s
Peak memory	38.2 GiB
Avg memory	30.6 GiB
CPU user	1146.1s
CPU sys	96.0s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	126.0s
Peak memory	41.8 GiB
Avg memory	30.2 GiB
CPU user	1144.0s
CPU sys	98.3s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-04-04T21:03:04Z

🤖 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 coalesce-push-decoder-ranges
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                      HEAD ┃               coalesce-push-decoder-ranges ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │               6.30 / 6.75 ±0.81 / 8.36 ms │                6.23 / 6.71 ±0.81 / 8.32 ms │     no change │
│ QQuery 2  │        112.07 / 141.89 ±24.16 / 165.06 ms │          110.57 / 111.72 ±0.97 / 113.50 ms │ +1.27x faster │
│ QQuery 3  │         109.87 / 111.11 ±0.81 / 112.34 ms │          108.72 / 109.44 ±0.45 / 110.10 ms │     no change │
│ QQuery 4  │      1125.21 / 1131.97 ±5.15 / 1140.64 ms │       1077.12 / 1094.26 ±9.70 / 1107.21 ms │     no change │
│ QQuery 5  │         190.92 / 198.83 ±4.62 / 204.43 ms │          190.90 / 195.83 ±3.07 / 200.25 ms │     no change │
│ QQuery 6  │        259.91 / 276.45 ±12.44 / 294.05 ms │          251.11 / 255.39 ±4.29 / 261.28 ms │ +1.08x faster │
│ QQuery 7  │        321.45 / 344.45 ±12.35 / 357.23 ms │          313.94 / 325.93 ±6.97 / 332.20 ms │ +1.06x faster │
│ QQuery 8  │         162.00 / 166.78 ±4.20 / 172.48 ms │          159.48 / 162.28 ±1.48 / 163.67 ms │     no change │
│ QQuery 9  │        221.75 / 245.56 ±14.56 / 265.98 ms │         216.87 / 247.95 ±17.44 / 267.59 ms │     no change │
│ QQuery 10 │         175.31 / 182.98 ±6.97 / 192.05 ms │          168.35 / 173.41 ±3.76 / 177.87 ms │ +1.06x faster │
│ QQuery 11 │        721.60 / 738.33 ±13.06 / 754.27 ms │          704.45 / 712.75 ±5.14 / 720.67 ms │     no change │
│ QQuery 12 │            39.51 / 40.77 ±0.81 / 41.73 ms │             37.09 / 39.00 ±1.79 / 41.98 ms │     no change │
│ QQuery 13 │         577.70 / 585.74 ±7.92 / 595.39 ms │          563.60 / 568.76 ±3.82 / 574.64 ms │     no change │
│ QQuery 14 │        899.04 / 926.05 ±14.12 / 938.52 ms │         891.93 / 904.66 ±11.09 / 922.97 ms │     no change │
│ QQuery 15 │            19.04 / 21.88 ±2.72 / 26.97 ms │             19.14 / 22.09 ±2.63 / 25.60 ms │     no change │
│ QQuery 16 │               7.01 / 7.33 ±0.19 / 7.50 ms │                6.87 / 7.44 ±0.57 / 8.48 ms │     no change │
│ QQuery 17 │         203.30 / 207.18 ±1.99 / 208.63 ms │          202.55 / 206.15 ±2.11 / 208.32 ms │     no change │
│ QQuery 18 │        412.85 / 472.78 ±31.37 / 499.23 ms │          473.57 / 477.51 ±2.53 / 480.35 ms │     no change │
│ QQuery 19 │         141.04 / 146.34 ±2.99 / 149.75 ms │          140.39 / 143.52 ±1.72 / 145.22 ms │     no change │
│ QQuery 20 │            15.63 / 16.43 ±0.73 / 17.38 ms │             15.50 / 16.73 ±1.17 / 18.77 ms │     no change │
│ QQuery 21 │            26.41 / 28.71 ±1.46 / 31.03 ms │             27.55 / 28.27 ±0.48 / 28.82 ms │     no change │
│ QQuery 22 │         507.00 / 517.18 ±8.09 / 531.28 ms │          512.08 / 522.70 ±7.50 / 532.12 ms │     no change │
│ QQuery 23 │      1248.41 / 1252.60 ±2.93 / 1256.38 ms │      1228.53 / 1253.07 ±18.80 / 1281.37 ms │     no change │
│ QQuery 24 │         118.36 / 124.08 ±3.40 / 129.05 ms │          120.77 / 124.66 ±2.67 / 127.64 ms │     no change │
│ QQuery 25 │         305.55 / 310.69 ±4.33 / 318.67 ms │          298.94 / 306.75 ±5.81 / 317.03 ms │     no change │
│ QQuery 26 │         143.55 / 151.32 ±5.17 / 157.30 ms │          138.82 / 151.47 ±8.89 / 164.93 ms │     no change │
│ QQuery 27 │               6.52 / 7.11 ±0.62 / 8.08 ms │                6.53 / 7.18 ±0.95 / 9.04 ms │     no change │
│ QQuery 28 │         219.52 / 227.29 ±4.90 / 232.57 ms │          222.85 / 226.78 ±2.93 / 230.93 ms │     no change │
│ QQuery 29 │         250.54 / 260.31 ±5.94 / 269.12 ms │          246.19 / 251.19 ±2.68 / 253.81 ms │     no change │
│ QQuery 30 │            56.70 / 59.63 ±2.47 / 63.99 ms │             57.24 / 59.98 ±1.78 / 61.93 ms │     no change │
│ QQuery 31 │         180.12 / 180.97 ±0.66 / 181.96 ms │          178.87 / 180.60 ±1.38 / 182.63 ms │     no change │
│ QQuery 32 │            52.67 / 54.89 ±1.35 / 56.72 ms │             52.94 / 55.07 ±1.78 / 57.24 ms │     no change │
│ QQuery 33 │         133.28 / 135.21 ±1.54 / 136.96 ms │          131.48 / 132.76 ±1.07 / 134.06 ms │     no change │
│ QQuery 34 │               7.03 / 7.44 ±0.70 / 8.85 ms │                6.99 / 7.47 ±0.53 / 8.44 ms │     no change │
│ QQuery 35 │         150.78 / 153.26 ±1.74 / 155.76 ms │          146.83 / 151.00 ±3.24 / 156.40 ms │     no change │
│ QQuery 36 │               6.38 / 6.61 ±0.19 / 6.86 ms │                6.33 / 6.47 ±0.20 / 6.86 ms │     no change │
│ QQuery 37 │               4.23 / 4.35 ±0.09 / 4.50 ms │                4.21 / 4.37 ±0.13 / 4.57 ms │     no change │
│ QQuery 38 │         109.87 / 111.91 ±1.53 / 113.28 ms │          103.98 / 108.44 ±3.15 / 113.72 ms │     no change │
│ QQuery 39 │         139.61 / 141.12 ±1.22 / 142.35 ms │          136.27 / 139.03 ±1.67 / 141.09 ms │     no change │
│ QQuery 40 │         137.85 / 145.56 ±9.25 / 163.70 ms │          129.89 / 140.76 ±8.32 / 154.99 ms │     no change │
│ QQuery 41 │            14.67 / 16.61 ±1.52 / 19.17 ms │             14.41 / 15.78 ±1.08 / 17.03 ms │     no change │
│ QQuery 42 │         116.40 / 117.83 ±1.15 / 119.34 ms │          106.91 / 112.16 ±4.10 / 116.24 ms │     no change │
│ QQuery 43 │               5.85 / 6.13 ±0.23 / 6.42 ms │                5.68 / 5.89 ±0.21 / 6.29 ms │     no change │
│ QQuery 44 │            11.20 / 11.58 ±0.42 / 12.34 ms │             10.95 / 12.53 ±1.41 / 14.53 ms │  1.08x slower │
│ QQuery 45 │            46.42 / 47.49 ±0.89 / 48.89 ms │             42.83 / 46.10 ±3.23 / 52.29 ms │     no change │
│ QQuery 46 │              8.29 / 8.91 ±0.64 / 10.08 ms │                8.23 / 8.71 ±0.51 / 9.39 ms │     no change │
│ QQuery 47 │         802.04 / 812.82 ±7.02 / 821.06 ms │         783.00 / 796.17 ±12.24 / 817.01 ms │     no change │
│ QQuery 48 │         493.91 / 501.44 ±6.23 / 510.86 ms │          481.28 / 483.85 ±1.83 / 486.59 ms │     no change │
│ QQuery 49 │         284.26 / 286.25 ±1.64 / 289.20 ms │          272.24 / 282.30 ±6.39 / 288.44 ms │     no change │
│ QQuery 50 │         543.03 / 554.11 ±7.21 / 565.77 ms │          521.32 / 536.89 ±9.60 / 546.94 ms │     no change │
│ QQuery 51 │         216.12 / 224.25 ±6.10 / 234.41 ms │          216.05 / 223.37 ±5.85 / 234.03 ms │     no change │
│ QQuery 52 │         108.50 / 114.46 ±4.13 / 119.07 ms │          107.63 / 112.31 ±2.71 / 115.03 ms │     no change │
│ QQuery 53 │         146.22 / 150.23 ±2.21 / 151.99 ms │          138.18 / 145.90 ±4.38 / 151.08 ms │     no change │
│ QQuery 54 │         129.95 / 131.92 ±1.81 / 134.22 ms │          126.52 / 128.96 ±1.78 / 131.20 ms │     no change │
│ QQuery 55 │         108.73 / 114.07 ±3.25 / 118.45 ms │          113.94 / 115.37 ±1.55 / 117.34 ms │     no change │
│ QQuery 56 │         133.25 / 135.15 ±1.97 / 138.68 ms │          127.67 / 132.10 ±2.33 / 134.27 ms │     no change │
│ QQuery 57 │         190.39 / 194.63 ±2.97 / 198.85 ms │          189.07 / 192.68 ±2.04 / 194.61 ms │     no change │
│ QQuery 58 │         371.02 / 379.97 ±7.37 / 390.29 ms │         366.52 / 379.48 ±10.00 / 392.28 ms │     no change │
│ QQuery 59 │         271.38 / 281.18 ±5.56 / 288.08 ms │          270.59 / 277.70 ±3.83 / 281.95 ms │     no change │
│ QQuery 60 │         137.71 / 139.72 ±1.74 / 142.78 ms │          136.00 / 138.81 ±1.81 / 141.35 ms │     no change │
│ QQuery 61 │            11.98 / 13.21 ±1.00 / 14.74 ms │             12.27 / 14.67 ±4.14 / 22.91 ms │  1.11x slower │
│ QQuery 62 │       949.71 / 988.95 ±58.88 / 1104.50 ms │        928.07 / 978.16 ±52.31 / 1071.35 ms │     no change │
│ QQuery 63 │         145.69 / 150.57 ±3.73 / 154.53 ms │          141.76 / 148.54 ±5.56 / 158.16 ms │     no change │
│ QQuery 64 │ 28012.40 / 29041.43 ±926.60 / 30665.55 ms │ 28255.25 / 29857.16 ±1524.41 / 32614.99 ms │     no change │
│ QQuery 65 │         357.49 / 369.41 ±9.39 / 381.89 ms │          358.39 / 373.15 ±7.98 / 382.39 ms │     no change │
│ QQuery 66 │         194.58 / 200.27 ±4.36 / 206.54 ms │          196.21 / 201.21 ±3.88 / 206.94 ms │     no change │
│ QQuery 67 │         510.90 / 517.74 ±8.04 / 532.78 ms │          503.77 / 517.02 ±9.89 / 531.59 ms │     no change │
│ QQuery 68 │             9.34 / 10.63 ±0.70 / 11.35 ms │               8.88 / 9.81 ±0.82 / 10.81 ms │ +1.08x faster │
│ QQuery 69 │         168.51 / 177.45 ±4.83 / 182.24 ms │          170.18 / 177.74 ±5.79 / 185.59 ms │     no change │
│ QQuery 70 │         432.41 / 437.99 ±5.06 / 446.49 ms │          430.99 / 442.42 ±5.85 / 447.18 ms │     no change │
│ QQuery 71 │         127.69 / 133.81 ±4.01 / 139.89 ms │          127.99 / 131.48 ±2.37 / 134.68 ms │     no change │
│ QQuery 72 │     1202.90 / 1221.41 ±11.29 / 1237.84 ms │       1191.54 / 1206.33 ±9.95 / 1220.89 ms │     no change │
│ QQuery 73 │               6.71 / 7.25 ±0.48 / 8.15 ms │                6.90 / 7.72 ±0.87 / 9.29 ms │  1.07x slower │
│ QQuery 74 │         502.97 / 510.85 ±4.75 / 515.56 ms │          505.48 / 508.47 ±2.21 / 512.21 ms │     no change │
│ QQuery 75 │         287.42 / 292.69 ±5.23 / 302.25 ms │          285.01 / 290.24 ±3.09 / 293.32 ms │     no change │
│ QQuery 76 │         279.32 / 285.46 ±5.39 / 295.09 ms │          273.84 / 281.19 ±6.21 / 292.03 ms │     no change │
│ QQuery 77 │         227.27 / 229.02 ±2.47 / 233.82 ms │          219.10 / 226.25 ±4.15 / 230.40 ms │     no change │
│ QQuery 78 │         307.32 / 313.22 ±4.77 / 321.15 ms │          304.14 / 309.54 ±4.63 / 316.96 ms │     no change │
│ QQuery 79 │         259.05 / 264.19 ±2.63 / 266.35 ms │          257.10 / 260.79 ±2.24 / 263.54 ms │     no change │
│ QQuery 80 │         319.55 / 327.31 ±3.97 / 330.68 ms │          323.36 / 325.40 ±1.82 / 328.57 ms │     no change │
│ QQuery 81 │            31.69 / 35.00 ±3.18 / 39.93 ms │             31.02 / 33.23 ±1.51 / 35.06 ms │ +1.05x faster │
│ QQuery 82 │         193.12 / 197.97 ±2.47 / 199.84 ms │          182.11 / 192.89 ±5.87 / 197.89 ms │     no change │
│ QQuery 83 │            46.02 / 47.20 ±1.36 / 49.81 ms │             45.26 / 47.04 ±1.47 / 49.40 ms │     no change │
│ QQuery 84 │            63.92 / 64.96 ±0.67 / 65.99 ms │             63.97 / 65.70 ±1.41 / 67.78 ms │     no change │
│ QQuery 85 │         277.19 / 289.36 ±7.57 / 300.97 ms │          274.78 / 283.99 ±7.09 / 293.79 ms │     no change │
│ QQuery 86 │            49.17 / 51.08 ±1.57 / 53.40 ms │             48.70 / 50.89 ±1.60 / 53.48 ms │     no change │
│ QQuery 87 │         103.24 / 107.97 ±4.81 / 116.76 ms │          101.13 / 109.02 ±4.23 / 112.61 ms │     no change │
│ QQuery 88 │         124.40 / 133.27 ±5.69 / 142.24 ms │          129.25 / 132.49 ±2.82 / 136.99 ms │     no change │
│ QQuery 89 │         149.67 / 158.97 ±6.78 / 166.39 ms │          149.90 / 157.87 ±4.12 / 161.15 ms │     no change │
│ QQuery 90 │            23.96 / 28.40 ±2.38 / 31.12 ms │             23.99 / 26.48 ±2.52 / 31.05 ms │ +1.07x faster │
│ QQuery 91 │          96.65 / 104.67 ±4.75 / 109.20 ms │          100.07 / 104.02 ±2.91 / 107.82 ms │     no change │
│ QQuery 92 │            53.65 / 54.91 ±1.33 / 57.30 ms │             50.85 / 54.31 ±2.44 / 57.32 ms │     no change │
│ QQuery 93 │         193.97 / 197.45 ±4.00 / 204.92 ms │          188.64 / 193.51 ±2.67 / 196.73 ms │     no change │
│ QQuery 94 │            64.01 / 65.88 ±1.67 / 68.65 ms │             60.68 / 62.90 ±1.65 / 65.23 ms │     no change │
│ QQuery 95 │         142.62 / 144.85 ±2.44 / 149.07 ms │          139.76 / 142.94 ±2.46 / 145.25 ms │     no change │
│ QQuery 96 │            80.20 / 88.16 ±3.99 / 90.51 ms │             88.10 / 88.92 ±0.93 / 90.51 ms │     no change │
│ QQuery 97 │         159.89 / 161.35 ±1.22 / 162.91 ms │          152.83 / 158.81 ±4.11 / 164.19 ms │     no change │
│ QQuery 98 │         123.11 / 124.49 ±1.14 / 126.40 ms │          123.09 / 124.89 ±1.91 / 127.74 ms │     no change │
│ QQuery 99 │  10806.31 / 10819.70 ±14.36 / 10846.50 ms │   10851.15 / 10871.02 ±17.30 / 10898.74 ms │     no change │
└───────────┴───────────────────────────────────────────┴────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                           ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                           │ 62447.08ms │
│ Total Time (coalesce-push-decoder-ranges)   │ 62954.80ms │
│ Average Time (HEAD)                         │   630.78ms │
│ Average Time (coalesce-push-decoder-ranges) │   635.91ms │
│ Queries Faster                              │          7 │
│ Queries Slower                              │          3 │
│ Queries with No Change                      │         89 │
│ Queries with Failure                        │          0 │
└─────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	312.7s
Peak memory	36.3 GiB
Avg memory	26.1 GiB
CPU user	610.7s
CPU sys	30.3s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	315.2s
Peak memory	36.7 GiB
Avg memory	25.9 GiB
CPU user	631.7s
CPU sys	27.5s
Peak spill	0 B

File an issue against this benchmark runner

github-actions Bot added core Core DataFusion crate datasource Changes to the datasource crate labels Apr 4, 2026

apache deleted a comment from adriangbot Apr 4, 2026

Dandandan marked this pull request as ready for review April 4, 2026 20:43

Dandandan closed this Apr 4, 2026

sesteves mentioned this pull request Aug 20, 2026

Filtered Parquet scans issue additional object-store requests after ParquetPushDecoder migration #24527

Open

Conversation

Dandandan commented Apr 4, 2026

Which issue does this PR close?

Rationale for this change

What changes are included in this PR?

Are these changes tested?

Are there any user-facing changes?

Uh oh!

Dandandan commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

Dandandan commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

Dandandan commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

adriangbot commented Apr 4, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants