feat: add BuildHasher variants for hash_utils

[xudong963]

Which issue does this PR close?

• Closes Add BuildHasher variants for hash_utils #21428.

Rationale for this change

datafusion_common::hash_utils currently exposes hashing helpers only for DataFusion's
RandomState. This makes the code harder to reuse from callers that need a custom
BuildHasher for integration or test scenarios.

What changes are included in this PR?

• Add with_hashes_with_hasher and create_hashes_with_hasher public APIs.
• Preserve the existing optimized RandomState path for the current with_hashes
  and create_hashes APIs.
• Extend the hash_utils tests to cover the new custom-hasher entry points.
• Gate the inequality assertion in the new custom-hasher test under
  force_hash_collisions so the feature-specific test configuration keeps passing.

Are these changes tested?

Yes.

• cargo test -p datafusion-common hash_utils
• cargo test -p datafusion-common --features force_hash_collisions hash_utils

Are there any user-facing changes?

Yes. datafusion-common now exposes with_hashes_with_hasher and
create_hashes_with_hasher for callers that need a custom BuildHasher.

[Dandandan]

run benchmark with_hashes

[adriangbot]

🤖 Criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4199049345-923-nb2ps 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 xudong963/upstream-pr46 (6a1cc02) to cdfade5 (merge-base) diff
BENCH_NAME=with_hashes
BENCH_COMMAND=cargo bench --features=parquet --bench with_hashes
BENCH_FILTER=
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Criterion 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

group                                        main                                   xudong963_upstream-pr46
-----                                        ----                                   -----------------------
dense_union: multiple, no nulls              1.00    239.1±0.25µs        ? ?/sec    1.20    286.4±1.29µs        ? ?/sec
dense_union: single, no nulls                1.00     80.1±0.09µs        ? ?/sec    1.20     95.8±0.24µs        ? ?/sec
dictionary_utf8_int32: multiple, no nulls    1.01     14.7±0.18µs        ? ?/sec    1.00     14.6±0.01µs        ? ?/sec
dictionary_utf8_int32: multiple, nulls       1.01     32.5±0.05µs        ? ?/sec    1.00     32.0±0.02µs        ? ?/sec
dictionary_utf8_int32: single, no nulls      1.00      4.6±0.01µs        ? ?/sec    1.01      4.7±0.08µs        ? ?/sec
dictionary_utf8_int32: single, nulls         1.01     10.9±0.03µs        ? ?/sec    1.00     10.8±0.00µs        ? ?/sec
int64: multiple, no nulls                    1.00     16.9±0.01µs        ? ?/sec    1.00     16.9±0.01µs        ? ?/sec
int64: multiple, nulls                       1.00     26.8±0.01µs        ? ?/sec    1.01     27.0±0.02µs        ? ?/sec
int64: single, no nulls                      1.00      4.6±0.01µs        ? ?/sec    1.00      4.5±0.00µs        ? ?/sec
int64: single, nulls                         1.00      8.7±0.02µs        ? ?/sec    1.02      8.9±0.01µs        ? ?/sec
large_utf8: multiple, no nulls               1.07    112.3±0.57µs        ? ?/sec    1.00    105.1±0.32µs        ? ?/sec
large_utf8: multiple, nulls                  1.13    158.9±0.52µs        ? ?/sec    1.00    141.3±1.71µs        ? ?/sec
large_utf8: single, no nulls                 1.00     26.5±0.09µs        ? ?/sec    1.37     36.2±0.41µs        ? ?/sec
large_utf8: single, nulls                    1.00     43.7±0.15µs        ? ?/sec    1.00     43.9±0.88µs        ? ?/sec
list_array: multiple, no nulls               1.00    152.3±0.30µs        ? ?/sec    1.00    153.0±0.48µs        ? ?/sec
list_array: multiple, nulls                  1.01    176.9±0.39µs        ? ?/sec    1.00    175.5±0.23µs        ? ?/sec
list_array: single, no nulls                 1.02     49.4±0.76µs        ? ?/sec    1.00     48.6±0.04µs        ? ?/sec
list_array: single, nulls                    1.01     59.3±0.15µs        ? ?/sec    1.00     58.8±0.07µs        ? ?/sec
list_array_sliced: 1/10 of 81920 rows        1.04     50.5±0.24µs        ? ?/sec    1.00     48.8±0.05µs        ? ?/sec
list_array_sliced: 1/2 of 16384 rows         1.00     49.0±0.05µs        ? ?/sec    1.01     49.5±0.08µs        ? ?/sec
list_array_sliced: 1/5 of 40960 rows         1.00     49.0±0.06µs        ? ?/sec    1.01     49.6±0.12µs        ? ?/sec
map_array: multiple, no nulls                1.00    243.7±1.01µs        ? ?/sec    1.00    242.8±0.25µs        ? ?/sec
map_array: multiple, nulls                   1.00    247.0±0.26µs        ? ?/sec    1.00    246.6±0.19µs        ? ?/sec
map_array: single, no nulls                  1.01     81.7±0.26µs        ? ?/sec    1.00     80.8±0.09µs        ? ?/sec
map_array: single, nulls                     1.00     82.8±0.05µs        ? ?/sec    1.00     82.4±0.11µs        ? ?/sec
map_array_sliced: 1/10 of 81920 rows         1.00     81.2±0.16µs        ? ?/sec    1.00     81.3±0.56µs        ? ?/sec
map_array_sliced: 1/2 of 16384 rows          1.00     80.8±0.06µs        ? ?/sec    1.00     81.1±0.06µs        ? ?/sec
map_array_sliced: 1/5 of 40960 rows          1.00     80.9±0.13µs        ? ?/sec    1.00     80.9±0.21µs        ? ?/sec
run_array_int32: multiple, no nulls          1.00      5.0±0.01µs        ? ?/sec    1.00      5.0±0.01µs        ? ?/sec
run_array_int32: multiple, nulls             1.00      6.7±0.01µs        ? ?/sec    1.00      6.7±0.01µs        ? ?/sec
run_array_int32: single, no nulls            1.00   1816.7±6.43ns        ? ?/sec    1.00   1813.2±3.31ns        ? ?/sec
run_array_int32: single, nulls               1.01      2.4±0.01µs        ? ?/sec    1.00      2.4±0.01µs        ? ?/sec
sparse_union: multiple, no nulls             1.00    261.1±0.51µs        ? ?/sec    1.78  463.9±149.35µs        ? ?/sec
sparse_union: single, no nulls               1.00     87.3±0.15µs        ? ?/sec    1.39    121.1±0.85µs        ? ?/sec
sparse_union_sliced: 1/10 of 81920 rows      1.00     87.5±0.22µs        ? ?/sec    1.38    120.7±0.83µs        ? ?/sec
sparse_union_sliced: 1/2 of 16384 rows       1.00     87.4±0.13µs        ? ?/sec    1.38    120.4±0.77µs        ? ?/sec
sparse_union_sliced: 1/5 of 40960 rows       1.00     87.6±0.17µs        ? ?/sec    1.38    120.5±0.94µs        ? ?/sec
struct_array: multiple, no nulls             1.00    185.4±1.26µs        ? ?/sec    1.01    188.2±0.52µs        ? ?/sec
struct_array: multiple, nulls                1.00    195.0±0.53µs        ? ?/sec    1.02    198.4±0.56µs        ? ?/sec
struct_array: single, no nulls               1.00     61.8±0.45µs        ? ?/sec    1.01     62.7±0.43µs        ? ?/sec
struct_array: single, nulls                  1.00     64.3±0.35µs        ? ?/sec    1.04     67.1±1.47µs        ? ?/sec
utf8: multiple, no nulls                     1.04   114.8±10.06µs        ? ?/sec    1.00    110.6±2.19µs        ? ?/sec
utf8: multiple, nulls                        1.02    142.6±1.13µs        ? ?/sec    1.00    139.3±0.46µs        ? ?/sec
utf8: single, no nulls                       1.02     34.6±1.67µs        ? ?/sec    1.00     33.8±0.13µs        ? ?/sec
utf8: single, nulls                          1.00     44.9±0.21µs        ? ?/sec    1.06     47.5±0.64µs        ? ?/sec
utf8_view (small): multiple, no nulls        1.00     19.0±0.01µs        ? ?/sec    1.00     19.0±0.01µs        ? ?/sec
utf8_view (small): multiple, nulls           1.00     29.4±0.03µs        ? ?/sec    1.01     29.7±0.02µs        ? ?/sec
utf8_view (small): single, no nulls          1.00      5.0±0.00µs        ? ?/sec    1.01      5.1±0.00µs        ? ?/sec
utf8_view (small): single, nulls             1.00      9.7±0.01µs        ? ?/sec    1.04     10.1±0.02µs        ? ?/sec
utf8_view: multiple, no nulls                1.00    109.0±0.32µs        ? ?/sec    1.01    110.5±0.64µs        ? ?/sec
utf8_view: multiple, nulls                   1.00    112.4±0.13µs        ? ?/sec    1.03    116.2±0.41µs        ? ?/sec
utf8_view: single, no nulls                  1.00     33.0±0.05µs        ? ?/sec    1.04     34.4±0.06µs        ? ?/sec
utf8_view: single, nulls                     1.00     35.7±0.05µs        ? ?/sec    1.04     37.2±0.06µs        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	529.0s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	664.3s
CPU sys	1.6s
Peak spill	0 B

branch

Metric	Value
Wall time	520.0s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	655.8s
CPU sys	0.7s
Peak spill	0 B

---

File an issue against this benchmark runner

[Dandandan]

run benchmark with_hashes

[Dandandan]

large_utf8: single, no nulls                 1.00     26.5±0.09µs        ? ?/sec    1.37     36.2±0.41µs        ? ?/sec
sparse_union: multiple, no nulls             1.00    261.1±0.51µs        ? ?/sec    1.78  463.9±149.35µs        ? ?/sec
sparse_union_sliced: 1/2 of 16384 rows       1.00     87.4±0.13µs        ? ?/sec    1.38    120.4±0.77µs        ? ?/sec

These might be regressions, I am rerunning once more.

[adriangbot]

🤖 Criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4204210371-928-m6h6p 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 xudong963/upstream-pr46 (6a1cc02) to cdfade5 (merge-base) diff
BENCH_NAME=with_hashes
BENCH_COMMAND=cargo bench --features=parquet --bench with_hashes
BENCH_FILTER=
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Criterion 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

group                                        main                                   xudong963_upstream-pr46
-----                                        ----                                   -----------------------
dense_union: multiple, no nulls              1.00    239.0±0.29µs        ? ?/sec    1.20    287.0±1.34µs        ? ?/sec
dense_union: single, no nulls                1.00     80.0±0.09µs        ? ?/sec    1.20     96.1±0.39µs        ? ?/sec
dictionary_utf8_int32: multiple, no nulls    1.00     14.6±0.01µs        ? ?/sec    1.00     14.6±0.01µs        ? ?/sec
dictionary_utf8_int32: multiple, nulls       1.02     32.6±0.03µs        ? ?/sec    1.00     32.0±0.03µs        ? ?/sec
dictionary_utf8_int32: single, no nulls      1.00      4.6±0.01µs        ? ?/sec    1.01      4.7±0.07µs        ? ?/sec
dictionary_utf8_int32: single, nulls         1.01     10.9±0.01µs        ? ?/sec    1.00     10.8±0.02µs        ? ?/sec
int64: multiple, no nulls                    1.00     16.9±0.02µs        ? ?/sec    1.00     16.9±0.01µs        ? ?/sec
int64: multiple, nulls                       1.00     26.8±0.01µs        ? ?/sec    1.01     27.1±0.03µs        ? ?/sec
int64: single, no nulls                      1.00      4.6±0.01µs        ? ?/sec    1.00      4.6±0.00µs        ? ?/sec
int64: single, nulls                         1.00      8.7±0.01µs        ? ?/sec    1.03      8.9±0.01µs        ? ?/sec
large_utf8: multiple, no nulls               1.06    111.9±0.43µs        ? ?/sec    1.00    105.5±0.31µs        ? ?/sec
large_utf8: multiple, nulls                  1.12    159.4±0.69µs        ? ?/sec    1.00    142.8±1.95µs        ? ?/sec
large_utf8: single, no nulls                 1.00     26.7±0.25µs        ? ?/sec    1.35     36.0±0.12µs        ? ?/sec
large_utf8: single, nulls                    1.00     44.6±0.71µs        ? ?/sec    1.00     44.4±0.86µs        ? ?/sec
list_array: multiple, no nulls               1.03    151.4±0.17µs        ? ?/sec    1.00    146.4±0.08µs        ? ?/sec
list_array: multiple, nulls                  1.00    175.7±0.20µs        ? ?/sec    1.01    176.6±0.08µs        ? ?/sec
list_array: single, no nulls                 1.00     48.8±0.04µs        ? ?/sec    1.05     51.2±0.06µs        ? ?/sec
list_array: single, nulls                    1.00     58.8±0.02µs        ? ?/sec    1.01     59.1±0.03µs        ? ?/sec
list_array_sliced: 1/10 of 81920 rows        1.04     51.4±0.31µs        ? ?/sec    1.00     49.3±1.00µs        ? ?/sec
list_array_sliced: 1/2 of 16384 rows         1.00     49.3±0.04µs        ? ?/sec    1.00     49.4±0.07µs        ? ?/sec
list_array_sliced: 1/5 of 40960 rows         1.03     50.9±0.06µs        ? ?/sec    1.00     49.3±0.07µs        ? ?/sec
map_array: multiple, no nulls                1.00    242.8±0.19µs        ? ?/sec    1.00    243.9±0.23µs        ? ?/sec
map_array: multiple, nulls                   1.00    248.1±0.72µs        ? ?/sec    1.00    247.7±1.35µs        ? ?/sec
map_array: single, no nulls                  1.00     81.3±0.04µs        ? ?/sec    1.01     82.1±0.19µs        ? ?/sec
map_array: single, nulls                     1.00     82.7±0.25µs        ? ?/sec    1.00     82.9±0.07µs        ? ?/sec
map_array_sliced: 1/10 of 81920 rows         1.00     81.0±0.06µs        ? ?/sec    1.00     81.3±0.04µs        ? ?/sec
map_array_sliced: 1/2 of 16384 rows          1.00     80.8±0.05µs        ? ?/sec    1.00     81.1±0.12µs        ? ?/sec
map_array_sliced: 1/5 of 40960 rows          1.00     81.5±0.07µs        ? ?/sec    1.00     81.8±0.21µs        ? ?/sec
run_array_int32: multiple, no nulls          1.01      5.0±0.01µs        ? ?/sec    1.00      4.9±0.01µs        ? ?/sec
run_array_int32: multiple, nulls             1.00      6.7±0.01µs        ? ?/sec    1.00      6.7±0.01µs        ? ?/sec
run_array_int32: single, no nulls            1.00   1821.5±6.62ns        ? ?/sec    1.00   1812.4±4.52ns        ? ?/sec
run_array_int32: single, nulls               1.01      2.4±0.01µs        ? ?/sec    1.00      2.4±0.01µs        ? ?/sec
sparse_union: multiple, no nulls             1.00    261.6±0.72µs        ? ?/sec    1.38    360.4±2.77µs        ? ?/sec
sparse_union: single, no nulls               1.00     87.4±0.11µs        ? ?/sec    1.38    120.3±1.06µs        ? ?/sec
sparse_union_sliced: 1/10 of 81920 rows      1.00     87.3±0.11µs        ? ?/sec    1.38    120.3±0.96µs        ? ?/sec
sparse_union_sliced: 1/2 of 16384 rows       1.00     87.3±0.15µs        ? ?/sec    1.38    120.6±0.78µs        ? ?/sec
sparse_union_sliced: 1/5 of 40960 rows       1.00     87.4±0.14µs        ? ?/sec    1.38    120.9±0.83µs        ? ?/sec
struct_array: multiple, no nulls             1.01    186.9±0.30µs        ? ?/sec    1.00    184.8±0.80µs        ? ?/sec
struct_array: multiple, nulls                1.00    196.0±1.14µs        ? ?/sec    1.00    196.1±0.28µs        ? ?/sec
struct_array: single, no nulls               1.00     62.6±0.11µs        ? ?/sec    1.00     62.5±0.15µs        ? ?/sec
struct_array: single, nulls                  1.02     65.7±0.33µs        ? ?/sec    1.00     64.5±0.09µs        ? ?/sec
utf8: multiple, no nulls                     1.02    111.4±0.74µs        ? ?/sec    1.00    109.3±1.77µs        ? ?/sec
utf8: multiple, nulls                        1.03    142.4±0.47µs        ? ?/sec    1.00    138.5±0.41µs        ? ?/sec
utf8: single, no nulls                       1.02     34.5±0.11µs        ? ?/sec    1.00     33.8±0.11µs        ? ?/sec
utf8: single, nulls                          1.00     45.7±0.82µs        ? ?/sec    1.01     46.4±0.72µs        ? ?/sec
utf8_view (small): multiple, no nulls        1.00     19.0±0.01µs        ? ?/sec    1.00     19.0±0.01µs        ? ?/sec
utf8_view (small): multiple, nulls           1.00     29.1±0.03µs        ? ?/sec    1.02     29.7±0.06µs        ? ?/sec
utf8_view (small): single, no nulls          1.00      5.1±0.00µs        ? ?/sec    1.00      5.0±0.01µs        ? ?/sec
utf8_view (small): single, nulls             1.00      9.7±0.01µs        ? ?/sec    1.04     10.1±0.01µs        ? ?/sec
utf8_view: multiple, no nulls                1.00    109.4±0.29µs        ? ?/sec    1.01    110.8±0.41µs        ? ?/sec
utf8_view: multiple, nulls                   1.00    112.4±0.20µs        ? ?/sec    1.03    115.5±0.26µs        ? ?/sec
utf8_view: single, no nulls                  1.00     33.0±0.04µs        ? ?/sec    1.04     34.4±0.04µs        ? ?/sec
utf8_view: single, nulls                     1.00     35.6±0.07µs        ? ?/sec    1.05     37.2±0.25µs        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	579.7s
Peak memory	3.4 GiB
Avg memory	3.4 GiB
CPU user	658.7s
CPU sys	1.5s
Peak spill	0 B

branch

Metric	Value
Wall time	518.6s
Peak memory	3.4 GiB
Avg memory	3.4 GiB
CPU user	654.6s
CPU sys	0.7s
Peak spill	0 B

---

File an issue against this benchmark runner

[xudong963]

run benchmark with_hashes

[adriangbot]

🤖 Criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4204691382-932-rps62 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 xudong963/upstream-pr46 (1dec571) to cdfade5 (merge-base) diff
BENCH_NAME=with_hashes
BENCH_COMMAND=cargo bench --features=parquet --bench with_hashes
BENCH_FILTER=
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Criterion 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

group                                        main                                   xudong963_upstream-pr46
-----                                        ----                                   -----------------------
dense_union: multiple, no nulls              1.00    239.0±0.49µs        ? ?/sec    1.20    287.1±0.64µs        ? ?/sec
dense_union: single, no nulls                1.00     79.9±0.11µs        ? ?/sec    1.20     96.1±0.22µs        ? ?/sec
dictionary_utf8_int32: multiple, no nulls    1.00     14.6±0.01µs        ? ?/sec    1.13     16.5±0.27µs        ? ?/sec
dictionary_utf8_int32: multiple, nulls       1.02     32.6±0.03µs        ? ?/sec    1.00     32.0±0.04µs        ? ?/sec
dictionary_utf8_int32: single, no nulls      1.01      4.6±0.01µs        ? ?/sec    1.00      4.6±0.01µs        ? ?/sec
dictionary_utf8_int32: single, nulls         1.00     10.9±0.01µs        ? ?/sec    1.00     10.9±0.03µs        ? ?/sec
int64: multiple, no nulls                    1.00     16.9±0.01µs        ? ?/sec    1.00     16.9±0.01µs        ? ?/sec
int64: multiple, nulls                       1.00     26.8±0.01µs        ? ?/sec    1.01     27.1±0.04µs        ? ?/sec
int64: single, no nulls                      1.00      4.6±0.00µs        ? ?/sec    1.00      4.6±0.01µs        ? ?/sec
int64: single, nulls                         1.00      8.7±0.01µs        ? ?/sec    1.02      8.9±0.01µs        ? ?/sec
large_utf8: multiple, no nulls               1.07    112.2±0.47µs        ? ?/sec    1.00    105.3±0.41µs        ? ?/sec
large_utf8: multiple, nulls                  1.12    158.9±0.52µs        ? ?/sec    1.00    141.6±1.83µs        ? ?/sec
large_utf8: single, no nulls                 1.00     26.3±0.13µs        ? ?/sec    1.36     35.8±0.50µs        ? ?/sec
large_utf8: single, nulls                    1.04     44.6±0.79µs        ? ?/sec    1.00     42.9±0.18µs        ? ?/sec
list_array: multiple, no nulls               1.00    148.3±0.10µs        ? ?/sec    1.01    149.3±0.29µs        ? ?/sec
list_array: multiple, nulls                  1.02    178.3±0.12µs        ? ?/sec    1.00    175.6±0.23µs        ? ?/sec
list_array: single, no nulls                 1.01     49.7±0.05µs        ? ?/sec    1.00     49.4±0.05µs        ? ?/sec
list_array: single, nulls                    1.02     59.7±0.13µs        ? ?/sec    1.00     58.6±0.03µs        ? ?/sec
list_array_sliced: 1/10 of 81920 rows        1.00     49.6±0.89µs        ? ?/sec    1.00     49.5±0.04µs        ? ?/sec
list_array_sliced: 1/2 of 16384 rows         1.00     48.9±0.02µs        ? ?/sec    1.01     49.6±0.02µs        ? ?/sec
list_array_sliced: 1/5 of 40960 rows         1.00     49.2±0.03µs        ? ?/sec    1.01     49.7±0.03µs        ? ?/sec
map_array: multiple, no nulls                1.01    242.4±0.18µs        ? ?/sec    1.00    240.7±0.17µs        ? ?/sec
map_array: multiple, nulls                   1.00    246.9±0.11µs        ? ?/sec    1.00    245.7±0.08µs        ? ?/sec
map_array: single, no nulls                  1.00     80.9±0.04µs        ? ?/sec    1.00     80.6±0.06µs        ? ?/sec
map_array: single, nulls                     1.01     82.6±0.04µs        ? ?/sec    1.00     82.1±0.05µs        ? ?/sec
map_array_sliced: 1/10 of 81920 rows         1.00     80.8±0.06µs        ? ?/sec    1.00     80.8±0.35µs        ? ?/sec
map_array_sliced: 1/2 of 16384 rows          1.00     80.7±0.05µs        ? ?/sec    1.00     80.6±0.06µs        ? ?/sec
map_array_sliced: 1/5 of 40960 rows          1.00     80.8±0.05µs        ? ?/sec    1.00     80.7±0.16µs        ? ?/sec
run_array_int32: multiple, no nulls          1.00      4.9±0.00µs        ? ?/sec    1.01      5.0±0.01µs        ? ?/sec
run_array_int32: multiple, nulls             1.01      6.7±0.02µs        ? ?/sec    1.00      6.7±0.02µs        ? ?/sec
run_array_int32: single, no nulls            1.00   1820.2±4.47ns        ? ?/sec    1.03   1874.9±3.52ns        ? ?/sec
run_array_int32: single, nulls               1.00      2.4±0.01µs        ? ?/sec    1.00      2.4±0.01µs        ? ?/sec
sparse_union: multiple, no nulls             1.00    260.9±0.43µs        ? ?/sec    1.38    360.0±2.39µs        ? ?/sec
sparse_union: single, no nulls               1.00     87.1±0.14µs        ? ?/sec    1.38    120.0±0.85µs        ? ?/sec
sparse_union_sliced: 1/10 of 81920 rows      1.00     87.1±0.12µs        ? ?/sec    1.38    120.7±0.77µs        ? ?/sec
sparse_union_sliced: 1/2 of 16384 rows       1.00     87.4±0.17µs        ? ?/sec    1.38    120.2±0.72µs        ? ?/sec
sparse_union_sliced: 1/5 of 40960 rows       1.00     87.4±0.11µs        ? ?/sec    1.38    120.9±1.11µs        ? ?/sec
struct_array: multiple, no nulls             1.00    184.4±1.14µs        ? ?/sec    1.02    187.2±0.48µs        ? ?/sec
struct_array: multiple, nulls                1.00    195.6±0.48µs        ? ?/sec    1.01    197.8±0.33µs        ? ?/sec
struct_array: single, no nulls               1.00     62.6±0.14µs        ? ?/sec    1.00     62.8±0.11µs        ? ?/sec
struct_array: single, nulls                  1.00     65.2±0.32µs        ? ?/sec    1.01     65.9±0.10µs        ? ?/sec
utf8: multiple, no nulls                     1.01    107.9±0.44µs        ? ?/sec    1.00    107.3±0.58µs        ? ?/sec
utf8: multiple, nulls                        1.03    142.7±0.78µs        ? ?/sec    1.00    138.8±0.57µs        ? ?/sec
utf8: single, no nulls                       1.00     33.8±0.57µs        ? ?/sec    1.00     33.8±0.07µs        ? ?/sec
utf8: single, nulls                          1.01     45.6±0.70µs        ? ?/sec    1.00     45.3±0.27µs        ? ?/sec
utf8_view (small): multiple, no nulls        1.00     19.0±0.01µs        ? ?/sec    1.00     19.0±0.01µs        ? ?/sec
utf8_view (small): multiple, nulls           1.00     29.3±0.02µs        ? ?/sec    1.01     29.7±0.03µs        ? ?/sec
utf8_view (small): single, no nulls          1.00      5.0±0.01µs        ? ?/sec    1.00      5.0±0.00µs        ? ?/sec
utf8_view (small): single, nulls             1.00      9.7±0.02µs        ? ?/sec    1.03     10.0±0.04µs        ? ?/sec
utf8_view: multiple, no nulls                1.00    109.4±0.55µs        ? ?/sec    1.00    108.9±0.66µs        ? ?/sec
utf8_view: multiple, nulls                   1.00    112.1±0.24µs        ? ?/sec    1.03    115.7±0.25µs        ? ?/sec
utf8_view: single, no nulls                  1.00     33.2±0.08µs        ? ?/sec    1.04     34.6±0.07µs        ? ?/sec
utf8_view: single, nulls                     1.00     35.6±0.06µs        ? ?/sec    1.05     37.3±0.09µs        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	524.3s
Peak memory	3.3 GiB
Avg memory	3.3 GiB
CPU user	660.3s
CPU sys	1.2s
Peak spill	0 B

branch

Metric	Value
Wall time	521.2s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	657.7s
CPU sys	0.8s
Peak spill	0 B

---

File an issue against this benchmark runner

[xudong963]

run benchmark with_hashes

[adriangbot]

🤖 Criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4212183047-981-2s69b 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 xudong963/upstream-pr46 (c022fb4) to cdfade5 (merge-base) diff
BENCH_NAME=with_hashes
BENCH_COMMAND=cargo bench --features=parquet --bench with_hashes
BENCH_FILTER=
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Criterion 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

group                                        main                                   xudong963_upstream-pr46
-----                                        ----                                   -----------------------
dense_union: multiple, no nulls              1.01    239.6±0.69µs        ? ?/sec    1.00    237.8±0.28µs        ? ?/sec
dense_union: single, no nulls                1.01     80.0±0.22µs        ? ?/sec    1.00     79.6±0.12µs        ? ?/sec
dictionary_utf8_int32: multiple, no nulls    1.00     14.7±0.05µs        ? ?/sec    1.12     16.5±0.01µs        ? ?/sec
dictionary_utf8_int32: multiple, nulls       1.02     32.6±0.05µs        ? ?/sec    1.00     32.1±0.02µs        ? ?/sec
dictionary_utf8_int32: single, no nulls      1.02      4.7±0.04µs        ? ?/sec    1.00      4.6±0.00µs        ? ?/sec
dictionary_utf8_int32: single, nulls         1.00     10.9±0.01µs        ? ?/sec    1.00     10.9±0.03µs        ? ?/sec
int64: multiple, no nulls                    1.00     16.9±0.01µs        ? ?/sec    1.00     16.9±0.01µs        ? ?/sec
int64: multiple, nulls                       1.00     26.9±0.06µs        ? ?/sec    1.00     26.8±0.01µs        ? ?/sec
int64: single, no nulls                      1.00      4.6±0.00µs        ? ?/sec    1.00      4.6±0.00µs        ? ?/sec
int64: single, nulls                         1.00      8.7±0.01µs        ? ?/sec    1.00      8.7±0.01µs        ? ?/sec
large_utf8: multiple, no nulls               1.00    112.0±0.44µs        ? ?/sec    1.14    127.2±0.43µs        ? ?/sec
large_utf8: multiple, nulls                  1.00    159.1±0.65µs        ? ?/sec    1.01    160.8±0.73µs        ? ?/sec
large_utf8: single, no nulls                 1.00     27.0±0.26µs        ? ?/sec    1.33     35.9±0.50µs        ? ?/sec
large_utf8: single, nulls                    1.04     44.9±0.74µs        ? ?/sec    1.00     43.1±0.19µs        ? ?/sec
list_array: multiple, no nulls               1.00    145.4±0.33µs        ? ?/sec    1.04    151.7±0.10µs        ? ?/sec
list_array: multiple, nulls                  1.00    175.5±0.15µs        ? ?/sec    1.00    175.8±0.12µs        ? ?/sec
list_array: single, no nulls                 1.00     48.8±0.13µs        ? ?/sec    1.00     48.6±0.05µs        ? ?/sec
list_array: single, nulls                    1.00     59.1±0.18µs        ? ?/sec    1.00     58.9±0.03µs        ? ?/sec
list_array_sliced: 1/10 of 81920 rows        1.00     49.0±0.04µs        ? ?/sec    1.05     51.3±0.06µs        ? ?/sec
list_array_sliced: 1/2 of 16384 rows         1.03     50.8±0.46µs        ? ?/sec    1.00     49.2±0.03µs        ? ?/sec
list_array_sliced: 1/5 of 40960 rows         1.00     49.4±0.61µs        ? ?/sec    1.00     49.2±0.03µs        ? ?/sec
map_array: multiple, no nulls                1.00    242.1±0.16µs        ? ?/sec    1.00    241.9±0.09µs        ? ?/sec
map_array: multiple, nulls                   1.00    249.1±1.72µs        ? ?/sec    1.00    249.0±0.13µs        ? ?/sec
map_array: single, no nulls                  1.00     80.9±0.07µs        ? ?/sec    1.00     80.9±0.04µs        ? ?/sec
map_array: single, nulls                     1.00     82.7±0.03µs        ? ?/sec    1.01     83.3±0.05µs        ? ?/sec
map_array_sliced: 1/10 of 81920 rows         1.00     80.9±0.07µs        ? ?/sec    1.00     80.8±0.05µs        ? ?/sec
map_array_sliced: 1/2 of 16384 rows          1.00     81.1±0.12µs        ? ?/sec    1.00     80.8±0.06µs        ? ?/sec
map_array_sliced: 1/5 of 40960 rows          1.00     80.8±0.05µs        ? ?/sec    1.00     81.1±0.04µs        ? ?/sec
run_array_int32: multiple, no nulls          1.00      5.0±0.01µs        ? ?/sec    1.00      4.9±0.01µs        ? ?/sec
run_array_int32: multiple, nulls             1.00      6.8±0.02µs        ? ?/sec    1.00      6.8±0.01µs        ? ?/sec
run_array_int32: single, no nulls            1.00   1810.6±3.71ns        ? ?/sec    1.00   1816.0±4.25ns        ? ?/sec
run_array_int32: single, nulls               1.00      2.4±0.01µs        ? ?/sec    1.00      2.4±0.01µs        ? ?/sec
sparse_union: multiple, no nulls             1.00    261.0±0.62µs        ? ?/sec    1.00    260.1±0.32µs        ? ?/sec
sparse_union: single, no nulls               1.00     87.2±0.16µs        ? ?/sec    1.00     87.0±0.12µs        ? ?/sec
sparse_union_sliced: 1/10 of 81920 rows      1.00     87.2±0.14µs        ? ?/sec    1.00     87.1±0.24µs        ? ?/sec
sparse_union_sliced: 1/2 of 16384 rows       1.00     87.5±0.13µs        ? ?/sec    1.00     87.2±0.14µs        ? ?/sec
sparse_union_sliced: 1/5 of 40960 rows       1.00     87.3±0.13µs        ? ?/sec    1.00     87.2±0.14µs        ? ?/sec
struct_array: multiple, no nulls             1.00    185.7±0.97µs        ? ?/sec    1.00    185.1±0.90µs        ? ?/sec
struct_array: multiple, nulls                1.00    195.6±0.28µs        ? ?/sec    1.00    196.0±0.40µs        ? ?/sec
struct_array: single, no nulls               1.00     62.1±0.37µs        ? ?/sec    1.00     62.1±0.27µs        ? ?/sec
struct_array: single, nulls                  1.00     65.1±0.14µs        ? ?/sec    1.00     65.0±0.10µs        ? ?/sec
utf8: multiple, no nulls                     1.00    107.8±0.32µs        ? ?/sec    1.01    109.0±1.30µs        ? ?/sec
utf8: multiple, nulls                        1.01    142.4±0.60µs        ? ?/sec    1.00    141.6±0.62µs        ? ?/sec
utf8: single, no nulls                       1.00     33.9±0.45µs        ? ?/sec    1.00     34.1±0.51µs        ? ?/sec
utf8: single, nulls                          1.03     45.9±0.49µs        ? ?/sec    1.00     44.4±0.19µs        ? ?/sec
utf8_view (small): multiple, no nulls        1.01     19.1±0.18µs        ? ?/sec    1.00     19.0±0.00µs        ? ?/sec
utf8_view (small): multiple, nulls           1.00     29.2±0.01µs        ? ?/sec    1.03     30.0±0.02µs        ? ?/sec
utf8_view (small): single, no nulls          1.00      5.0±0.00µs        ? ?/sec    1.00      5.0±0.00µs        ? ?/sec
utf8_view (small): single, nulls             1.00      9.6±0.01µs        ? ?/sec    1.00      9.7±0.01µs        ? ?/sec
utf8_view: multiple, no nulls                1.00    108.9±0.31µs        ? ?/sec    1.01    109.8±0.35µs        ? ?/sec
utf8_view: multiple, nulls                   1.01    112.3±0.20µs        ? ?/sec    1.00    111.6±0.16µs        ? ?/sec
utf8_view: single, no nulls                  1.00     33.0±0.07µs        ? ?/sec    1.00     33.0±0.07µs        ? ?/sec
utf8_view: single, nulls                     1.00     35.7±0.06µs        ? ?/sec    1.00     35.6±0.06µs        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	523.9s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	659.3s
CPU sys	1.3s
Peak spill	0 B

branch

Metric	Value
Wall time	529.7s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	666.3s
CPU sys	0.7s
Peak spill	0 B

---

File an issue against this benchmark runner

[xudong963]

run benchmark with_hashes

[adriangbot]

🤖 Criterion benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4212630263-985-7nb4f 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 xudong963/upstream-pr46 (501b223) to cdfade5 (merge-base) diff
BENCH_NAME=with_hashes
BENCH_COMMAND=cargo bench --features=parquet --bench with_hashes
BENCH_FILTER=
Results will be posted here when complete

---

File an issue against this benchmark runner

[adriangbot]

🤖 Criterion 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

group                                        main                                   xudong963_upstream-pr46
-----                                        ----                                   -----------------------
dense_union: multiple, no nulls              1.00    239.1±0.52µs        ? ?/sec    1.00    238.1±0.30µs        ? ?/sec
dense_union: single, no nulls                1.01     80.0±0.10µs        ? ?/sec    1.00     79.6±0.10µs        ? ?/sec
dictionary_utf8_int32: multiple, no nulls    1.00     14.6±0.01µs        ? ?/sec    1.13     16.4±0.01µs        ? ?/sec
dictionary_utf8_int32: multiple, nulls       1.02     32.7±0.03µs        ? ?/sec    1.00     32.2±0.02µs        ? ?/sec
dictionary_utf8_int32: single, no nulls      1.01      4.7±0.07µs        ? ?/sec    1.00      4.6±0.01µs        ? ?/sec
dictionary_utf8_int32: single, nulls         1.00     10.9±0.01µs        ? ?/sec    1.00     10.9±0.01µs        ? ?/sec
int64: multiple, no nulls                    1.00     16.9±0.01µs        ? ?/sec    1.00     16.9±0.01µs        ? ?/sec
int64: multiple, nulls                       1.00     26.8±0.01µs        ? ?/sec    1.00     26.8±0.01µs        ? ?/sec
int64: single, no nulls                      1.00      4.6±0.04µs        ? ?/sec    1.00      4.6±0.01µs        ? ?/sec
int64: single, nulls                         1.00      8.7±0.02µs        ? ?/sec    1.00      8.8±0.00µs        ? ?/sec
large_utf8: multiple, no nulls               1.00    111.9±0.44µs        ? ?/sec    1.14    127.2±0.49µs        ? ?/sec
large_utf8: multiple, nulls                  1.00    159.5±0.73µs        ? ?/sec    1.00    160.2±0.74µs        ? ?/sec
large_utf8: single, no nulls                 1.00     26.7±0.28µs        ? ?/sec    1.36     36.4±0.11µs        ? ?/sec
large_utf8: single, nulls                    1.01     44.1±0.24µs        ? ?/sec    1.00     43.9±0.64µs        ? ?/sec
list_array: multiple, no nulls               1.00    148.4±2.29µs        ? ?/sec    1.02    150.9±0.25µs        ? ?/sec
list_array: multiple, nulls                  1.00    175.8±0.13µs        ? ?/sec    1.00    175.8±0.08µs        ? ?/sec
list_array: single, no nulls                 1.01     49.2±0.09µs        ? ?/sec    1.00     48.6±0.09µs        ? ?/sec
list_array: single, nulls                    1.00     58.8±0.04µs        ? ?/sec    1.00     58.9±0.05µs        ? ?/sec
list_array_sliced: 1/10 of 81920 rows        1.00     48.7±0.04µs        ? ?/sec    1.03     50.4±0.08µs        ? ?/sec
list_array_sliced: 1/2 of 16384 rows         1.00     49.2±0.05µs        ? ?/sec    1.03     50.8±0.14µs        ? ?/sec
list_array_sliced: 1/5 of 40960 rows         1.00     49.2±0.08µs        ? ?/sec    1.04     51.3±0.03µs        ? ?/sec
map_array: multiple, no nulls                1.00    241.0±0.14µs        ? ?/sec    1.01    242.3±0.12µs        ? ?/sec
map_array: multiple, nulls                   1.00    246.0±0.46µs        ? ?/sec    1.01    249.2±0.43µs        ? ?/sec
map_array: single, no nulls                  1.00     80.7±0.12µs        ? ?/sec    1.00     80.9±0.08µs        ? ?/sec
map_array: single, nulls                     1.00     82.7±0.13µs        ? ?/sec    1.01     83.4±0.04µs        ? ?/sec
map_array_sliced: 1/10 of 81920 rows         1.00     80.8±0.08µs        ? ?/sec    1.00     81.1±0.07µs        ? ?/sec
map_array_sliced: 1/2 of 16384 rows          1.00     80.9±0.06µs        ? ?/sec    1.00     81.0±0.14µs        ? ?/sec
map_array_sliced: 1/5 of 40960 rows          1.00     80.9±0.05µs        ? ?/sec    1.00     80.7±0.08µs        ? ?/sec
run_array_int32: multiple, no nulls          1.01      5.0±0.01µs        ? ?/sec    1.00      4.9±0.01µs        ? ?/sec
run_array_int32: multiple, nulls             1.00      6.7±0.01µs        ? ?/sec    1.00      6.7±0.01µs        ? ?/sec
run_array_int32: single, no nulls            1.00   1815.4±4.60ns        ? ?/sec    1.00   1815.9±4.69ns        ? ?/sec
run_array_int32: single, nulls               1.00      2.4±0.00µs        ? ?/sec    1.00      2.4±0.00µs        ? ?/sec
sparse_union: multiple, no nulls             1.00    261.1±0.35µs        ? ?/sec    1.00    260.2±0.63µs        ? ?/sec
sparse_union: single, no nulls               1.01     87.4±0.17µs        ? ?/sec    1.00     86.9±0.15µs        ? ?/sec
sparse_union_sliced: 1/10 of 81920 rows      1.01     87.6±0.15µs        ? ?/sec    1.00     87.0±0.16µs        ? ?/sec
sparse_union_sliced: 1/2 of 16384 rows       1.01     87.7±0.12µs        ? ?/sec    1.00     87.1±0.13µs        ? ?/sec
sparse_union_sliced: 1/5 of 40960 rows       1.01     87.7±0.12µs        ? ?/sec    1.00     87.0±0.14µs        ? ?/sec
struct_array: multiple, no nulls             1.01    186.9±0.35µs        ? ?/sec    1.00    184.3±1.31µs        ? ?/sec
struct_array: multiple, nulls                1.00    194.4±0.99µs        ? ?/sec    1.00    193.7±0.92µs        ? ?/sec
struct_array: single, no nulls               1.01     62.4±0.11µs        ? ?/sec    1.00     61.8±0.56µs        ? ?/sec
struct_array: single, nulls                  1.01     65.4±0.16µs        ? ?/sec    1.00     64.9±0.34µs        ? ?/sec
utf8: multiple, no nulls                     1.00    108.3±0.57µs        ? ?/sec    1.00    108.8±1.38µs        ? ?/sec
utf8: multiple, nulls                        1.00    142.9±0.73µs        ? ?/sec    1.00    142.3±0.37µs        ? ?/sec
utf8: single, no nulls                       1.00     33.8±0.48µs        ? ?/sec    1.01     34.1±0.14µs        ? ?/sec
utf8: single, nulls                          1.00     45.3±0.33µs        ? ?/sec    1.00     45.4±0.53µs        ? ?/sec
utf8_view (small): multiple, no nulls        1.00     19.0±0.01µs        ? ?/sec    1.00     19.0±0.01µs        ? ?/sec
utf8_view (small): multiple, nulls           1.00     29.1±0.04µs        ? ?/sec    1.03     30.1±0.15µs        ? ?/sec
utf8_view (small): single, no nulls          1.00      5.0±0.00µs        ? ?/sec    1.00      5.0±0.00µs        ? ?/sec
utf8_view (small): single, nulls             1.00      9.6±0.01µs        ? ?/sec    1.01      9.7±0.01µs        ? ?/sec
utf8_view: multiple, no nulls                1.00    109.1±0.22µs        ? ?/sec    1.00    109.4±0.21µs        ? ?/sec
utf8_view: multiple, nulls                   1.00    112.3±0.19µs        ? ?/sec    1.00    112.1±0.25µs        ? ?/sec
utf8_view: single, no nulls                  1.00     33.0±0.04µs        ? ?/sec    1.00     32.9±0.06µs        ? ?/sec
utf8_view: single, nulls                     1.00     35.6±0.06µs        ? ?/sec    1.00     35.7±0.08µs        ? ?/sec

Resource Usage

base (merge-base)

Metric	Value
Wall time	524.0s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	659.9s
CPU sys	1.2s
Peak spill	0 B

branch

Metric	Value
Wall time	523.9s
Peak memory	3.4 GiB
Avg memory	3.3 GiB
CPU user	660.3s
CPU sys	0.7s
Peak spill	0 B

---

File an issue against this benchmark runner