BE-513: HashQL: Rework dynamic aggregate size estimation by indietyp · Pull Request #8697 · hashintel/hash

indietyp · 2026-05-04T14:23:48Z

🌟 What is the purpose of this PR?

The size estimation analysis previously treated all aggregate kinds (structs, tuples, lists, dicts, closures) identically — summing operand footprints and accumulating cardinality as if every aggregate were a flat collection. This was incorrect: a struct or tuple is a single composite value with cardinality 1, while a list or dict is a true collection whose cardinality equals its element count.

This PR introduces type-aware aggregate footprint evaluation. Structs and tuples now correctly report cardinality 1 with units equal to the sum of their fields' materialized sizes. Lists report per-element units (joined across elements) with cardinality equal to the element count. Dicts compute per-pair units (key + value combined) with cardinality equal to the pair count. Closures combine their function pointer and environment footprints into a single scalar value.

To support this, a materialize() method is introduced on Footprint that collapses a footprint's (units, cardinality) pair into a single total information estimate. This is needed when a value with its own cardinality (e.g. a list) is embedded as a field of a composite type — the field's contribution to the parent's units must account for the full information content of the nested value, not just its per-element size.

🔍 What does this change?

Replaces the single generic RValue::Aggregate handler in eval_rvalue with a dedicated eval_rvalue_aggregate method that dispatches on AggregateKind.
Struct/Tuple: sums the materialize()d footprints of all operands and sets cardinality to 1.
List: joins per-element materialized units and sets cardinality to the literal element count.
Dict: joins per-pair materialized units (key + value combined via saturating_mul_add) and sets cardinality to the pair count.
Closure: combines function pointer and environment footprints into a single scalar (cardinality 1).
Opaque: retains the previous behaviour of summing raw footprints.
Adds Footprint::materialize() which multiplies units by cardinality to produce a total information estimate, with case-specific handling for constant×constant (exact), affine units×constant cardinality (scale coefficients by cardinality upper bound), and affine×affine (element-wise coefficient multiplication as a linear under-approximation).
Adds Footprint::one(units) constructor for footprints with cardinality exactly 1.
Adds Estimate::saturating_coeff_mul for element-wise coefficient multiplication between two estimates.
Adds InformationRange::saturating_mul_cardinality for range-level multiplication of information by cardinality, saturating to unbounded on overflow.
Adds Eval::into_footprint as a consuming counterpart to Eval::as_ref.
Fixes the snapshot for struct_aggregate_sums_operands and tuple_aggregate_sums_operands, which previously reported cardinality 2 for a two-field struct/tuple; both now correctly report cardinality 1.

🛡 What tests cover this?

New integration tests: list_aggregate_per_element_units, dict_aggregate_per_pair_units, tuple_many_fields_cardinality_one, and struct_materializes_list_parameter, each with corresponding snapshot files.
New unit tests in footprint.rs covering all four materialize() branches: scalar identity, constant×constant, affine units×constant cardinality, constant units×affine cardinality, and both-affine same-parameter.
New unit tests in range.rs covering saturating_mul_cardinality: exact multiplication, identity by 1, empty inputs, unbounded cardinality, and overflow to unbounded.

❓ How to test this?

Run cargo test -p hashql-mir and confirm all tests pass.
Review the new and updated snapshots under libs/@local/hashql/mir/tests/ui/pass/size-estimation/ to verify the reported units and cardinality values match the expected semantics for each aggregate kind.

vercel · 2026-05-04T14:23:53Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
hash	Ready	Preview, Comment	May 4, 2026 3:33pm

2 Skipped Deployments

Project	Deployment	Actions	Updated (UTC)
hashdotdesign-tokens	Ignored	Preview	May 4, 2026 3:33pm
petrinaut	Skipped		May 4, 2026 3:33pm

cursor · 2026-05-04T14:23:57Z

PR Summary

Medium Risk
Changes core size-estimation semantics for aggregates (units/cardinality computation and affine handling), which can affect downstream planning/optimization behavior, though it’s covered by expanded unit + snapshot tests.

Overview
Reworks dynamic size estimation for RValue::Aggregate to be aggregate-kind aware instead of treating all aggregates as flat collections: structs/tuples now return cardinality 1 with units equal to the sum of each field’s materialized size; lists return joined per-element units with cardinality = element count; dicts return joined per-pair (key+value) units with cardinality = pair count; closures combine fn ptr + env into a single scalar footprint; opaque aggregates keep the prior “sum raw footprints” behavior.

Adds Footprint::materialize() (and helpers like Footprint::one, Estimate::saturating_coeff_mul, InformationRange::saturating_mul_cardinality, and Eval::into_footprint) to correctly collapse nested (units, cardinality) into total information when embedding collections inside composites. Updates/extends tests and snapshots to lock in the corrected cardinality semantics and new list/dict/materialization behavior.

^{Reviewed by Cursor Bugbot for commit 598d416. Bugbot is set up for automated code reviews on this repo. Configure here.}

indietyp · 2026-05-04T14:24:09Z

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

This stack of pull requests is managed by Graphite. Learn more about stacking.

augmentcode · 2026-05-04T14:27:57Z

🤖 Augment PR Summary

Summary: Reworks HashQL MIR dynamic aggregate size estimation to be type-aware, fixing incorrect cardinality behavior that previously treated all aggregates like flat collections.

Changes:

Adds `eval_rvalue_aggregate()` to dispatch aggregate evaluation by `AggregateKind` instead of generic operand summation.
Struct/Tuple: materialize each field and sum into `units`, with `cardinality = 1`.
List: joins per-element materialized `units`, with `cardinality` set to the literal element count.
Dict: joins per-pair (key + value) materialized `units`, with `cardinality` set to the pair count.
Closure: combines function pointer and environment footprints into a single scalar footprint.
Introduces `Footprint::materialize()` (+ `Footprint::one`, `Eval::into_footprint`) to collapse `(units, cardinality)` into total information when embedding nested values.
Adds supporting math helpers (`Estimate::saturating_coeff_mul`, `InformationRange::saturating_mul_cardinality`) plus new/updated tests and snapshots for the new semantics.

Why: Ensures composite values (structs/tuples/closures) aren’t mis-modeled as collections, and nested collections contribute their full information content when used inside composites.

_{🤖 Was this summary useful? React with 👍 or 👎}

augmentcode

Review completed. 1 suggestion posted.

Comment augment review to trigger a new review at any time.

codspeed-hq · 2026-05-04T14:29:56Z

Merging this PR will not alter performance

✅ 24 untouched benchmarks
⏩ 56 skipped benchmarks¹

_{Comparing bm/be-513-hashql-size-estimation-aggregate-cardinality-is-wrong-for (598d416) with bm/be-512-hashql-switchint-allow-cross-backend-transitions (5eeec53)}

56 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

codecov · 2026-05-04T14:31:44Z

Codecov Report

❌ Patch coverage is 91.07143% with 30 lines in your changes missing coverage. Please review.
✅ Project coverage is 66.31%. Comparing base (5eeec53) to head (598d416).

Files with missing lines	Patch %	Lines
...l/mir/src/pass/analysis/size_estimation/dynamic.rs	73.49%	22 Missing ⚠️
...mir/src/pass/analysis/size_estimation/footprint.rs	92.15%	4 Missing and 4 partials ⚠️

Additional details and impacted files

@@                                      Coverage Diff                                       @@
##           bm/be-512-hashql-switchint-allow-cross-backend-transitions    #8697      +/-   ##
==============================================================================================
+ Coverage                                                       66.19%   66.31%   +0.11%     
==============================================================================================
  Files                                                             932      932              
  Lines                                                           84621    84949     +328     
  Branches                                                         4461     4468       +7     
==============================================================================================
+ Hits                                                            56017    56331     +314     
- Misses                                                          28056    28067      +11     
- Partials                                                          548      551       +3

Flag	Coverage Δ
apps.hash-ai-worker-ts	`1.41% <ø> (ø)`
apps.hash-api	`0.00% <ø> (ø)`
local.hash-backend-utils	`0.00% <ø> (ø)`
local.hash-graph-sdk	`9.63% <ø> (ø)`
local.hash-isomorphic-utils	`0.00% <ø> (ø)`
rust.hash-graph-api	`2.52% <ø> (ø)`
rust.hashql-compiletest	`28.26% <ø> (ø)`
rust.hashql-eval	`79.70% <ø> (ø)`
rust.hashql-mir	`91.88% <91.07%> (+0.04%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

cursor

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 598d416. Configure here.}

cursor · 2026-05-04T15:36:51Z

+                    *coefficient = coefficient.saturating_mul(scale);
+                }
+
+                Estimate::Affine(result)


Empty cardinality treated as unbounded in materialize

Low Severity

In materialize(), the Affine × Constant arm calls cardinality.inclusive_max(), which returns None for both Cardinality::empty() (max = Excluded(0), since 0u32.checked_sub(1) is None) and Cardinality::full() (max = Unbounded). When None is returned, the code assumes unbounded cardinality and returns InformationRange::full(). This is wrong for empty cardinality — the result should be empty/zero, not full. The Constant × Constant arm handles this correctly because saturating_mul_cardinality explicitly checks is_empty() first. A missing cardinality.is_empty() guard before the inclusive_max() check causes the inconsistency.

Additional Locations (1)

libs/@local/hashql/mir/src/pass/analysis/size_estimation/footprint.rs#L281-L284

^{Reviewed by Cursor Bugbot for commit 598d416. Configure here.}

github-actions · 2026-05-04T16:47:41Z

Benchmark results

@rust/hash-graph-benches – Integrations

policy_resolution_large

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 2002	$$25.9 \mathrm{ms} \pm 137 \mathrm{μs}\left({\color{gray}1.37 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$3.55 \mathrm{ms} \pm 20.0 \mathrm{μs}\left({\color{gray}1.32 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 1001	$$12.5 \mathrm{ms} \pm 92.0 \mathrm{μs}\left({\color{gray}2.80 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 3314	$$42.6 \mathrm{ms} \pm 322 \mathrm{μs}\left({\color{gray}-0.602 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$15.4 \mathrm{ms} \pm 106 \mathrm{μs}\left({\color{gray}1.58 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 1526	$$23.6 \mathrm{ms} \pm 210 \mathrm{μs}\left({\color{gray}0.861 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 2078	$$26.3 \mathrm{ms} \pm 185 \mathrm{μs}\left({\color{gray}0.051 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.88 \mathrm{ms} \pm 22.3 \mathrm{μs}\left({\color{gray}1.48 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 1033	$$13.1 \mathrm{ms} \pm 77.2 \mathrm{μs}\left({\color{gray}-0.530 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_medium

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 102	$$3.70 \mathrm{ms} \pm 19.5 \mathrm{μs}\left({\color{gray}0.485 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$3.00 \mathrm{ms} \pm 17.1 \mathrm{μs}\left({\color{gray}0.759 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 51	$$3.32 \mathrm{ms} \pm 16.9 \mathrm{μs}\left({\color{gray}-0.491 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 269	$$5.11 \mathrm{ms} \pm 34.9 \mathrm{μs}\left({\color{gray}1.35 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$3.50 \mathrm{ms} \pm 17.7 \mathrm{μs}\left({\color{gray}-1.393 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 107	$$4.07 \mathrm{ms} \pm 29.7 \mathrm{μs}\left({\color{gray}0.508 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 133	$$4.30 \mathrm{ms} \pm 25.6 \mathrm{μs}\left({\color{gray}0.266 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.42 \mathrm{ms} \pm 15.3 \mathrm{μs}\left({\color{gray}-0.858 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 63	$$4.01 \mathrm{ms} \pm 23.9 \mathrm{μs}\left({\color{gray}-0.184 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_none

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 2	$$2.65 \mathrm{ms} \pm 11.2 \mathrm{μs}\left({\color{gray}2.85 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$2.52 \mathrm{ms} \pm 15.0 \mathrm{μs}\left({\color{gray}0.018 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 1	$$2.66 \mathrm{ms} \pm 13.0 \mathrm{μs}\left({\color{gray}3.79 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 8	$$2.93 \mathrm{ms} \pm 13.2 \mathrm{μs}\left({\color{gray}2.95 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$2.71 \mathrm{ms} \pm 13.0 \mathrm{μs}\left({\color{gray}3.41 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 3	$$2.93 \mathrm{ms} \pm 14.0 \mathrm{μs}\left({\color{gray}3.45 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_small

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 52	$$3.04 \mathrm{ms} \pm 19.5 \mathrm{μs}\left({\color{gray}2.41 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$2.71 \mathrm{ms} \pm 11.9 \mathrm{μs}\left({\color{gray}1.11 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 25	$$2.97 \mathrm{ms} \pm 17.8 \mathrm{μs}\left({\color{gray}-0.175 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 94	$$3.37 \mathrm{ms} \pm 20.0 \mathrm{μs}\left({\color{gray}1.01 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$3.00 \mathrm{ms} \pm 20.6 \mathrm{μs}\left({\color{gray}1.56 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 26	$$3.26 \mathrm{ms} \pm 16.8 \mathrm{μs}\left({\color{gray}0.775 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 66	$$3.28 \mathrm{ms} \pm 15.8 \mathrm{μs}\left({\color{gray}-0.266 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$2.96 \mathrm{ms} \pm 17.2 \mathrm{μs}\left({\color{gray}2.44 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 29	$$3.28 \mathrm{ms} \pm 17.9 \mathrm{μs}\left({\color{gray}-0.217 \mathrm{\%}}\right) $$	Flame Graph

read_scaling_complete

Function	Value	Mean	Flame graphs
entity_by_id;one_depth	1 entities	$$54.3 \mathrm{ms} \pm 372 \mathrm{μs}\left({\color{gray}-0.338 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	10 entities	$$45.9 \mathrm{ms} \pm 210 \mathrm{μs}\left({\color{gray}-0.261 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	25 entities	$$50.8 \mathrm{ms} \pm 215 \mathrm{μs}\left({\color{gray}1.35 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	5 entities	$$49.1 \mathrm{ms} \pm 1.06 \mathrm{ms}\left({\color{red}11.4 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	50 entities	$$62.5 \mathrm{ms} \pm 329 \mathrm{μs}\left({\color{gray}-0.253 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	1 entities	$$61.5 \mathrm{ms} \pm 299 \mathrm{μs}\left({\color{gray}-0.306 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	10 entities	$$55.6 \mathrm{ms} \pm 253 \mathrm{μs}\left({\color{gray}-0.289 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	25 entities	$$102 \mathrm{ms} \pm 459 \mathrm{μs}\left({\color{gray}0.893 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	5 entities	$$46.4 \mathrm{ms} \pm 280 \mathrm{μs}\left({\color{gray}0.465 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	50 entities	$$268 \mathrm{ms} \pm 784 \mathrm{μs}\left({\color{lightgreen}-6.953 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	1 entities	$$19.4 \mathrm{ms} \pm 92.2 \mathrm{μs}\left({\color{gray}1.19 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	10 entities	$$19.9 \mathrm{ms} \pm 101 \mathrm{μs}\left({\color{gray}-0.012 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	25 entities	$$20.3 \mathrm{ms} \pm 99.9 \mathrm{μs}\left({\color{gray}-0.191 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	5 entities	$$19.3 \mathrm{ms} \pm 90.7 \mathrm{μs}\left({\color{gray}0.313 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	50 entities	$$25.4 \mathrm{ms} \pm 126 \mathrm{μs}\left({\color{gray}0.955 \mathrm{\%}}\right) $$	Flame Graph

read_scaling_linkless

Function	Value	Mean	Flame graphs
entity_by_id	1 entities	$$19.3 \mathrm{ms} \pm 91.8 \mathrm{μs}\left({\color{gray}1.25 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	10 entities	$$19.1 \mathrm{ms} \pm 86.7 \mathrm{μs}\left({\color{gray}-0.857 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	100 entities	$$18.5 \mathrm{ms} \pm 87.8 \mathrm{μs}\left({\color{lightgreen}-5.355 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	1000 entities	$$20.0 \mathrm{ms} \pm 92.9 \mathrm{μs}\left({\color{gray}-3.271 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	10000 entities	$$26.7 \mathrm{ms} \pm 185 \mathrm{μs}\left({\color{gray}0.293 \mathrm{\%}}\right) $$	Flame Graph

representative_read_entity

Function	Value	Mean	Flame graphs
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/block/v/1`	$$33.7 \mathrm{ms} \pm 266 \mathrm{μs}\left({\color{gray}1.60 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/book/v/1`	$$34.9 \mathrm{ms} \pm 240 \mathrm{μs}\left({\color{gray}3.47 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/building/v/1`	$$33.7 \mathrm{ms} \pm 270 \mathrm{μs}\left({\color{gray}-0.437 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/organization/v/1`	$$34.5 \mathrm{ms} \pm 281 \mathrm{μs}\left({\color{gray}-0.176 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/page/v/2`	$$33.9 \mathrm{ms} \pm 254 \mathrm{μs}\left({\color{gray}0.822 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/person/v/1`	$$34.6 \mathrm{ms} \pm 252 \mathrm{μs}\left({\color{gray}1.90 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/playlist/v/1`	$$33.9 \mathrm{ms} \pm 261 \mathrm{μs}\left({\color{gray}-1.389 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/song/v/1`	$$34.2 \mathrm{ms} \pm 279 \mathrm{μs}\left({\color{gray}2.08 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/uk-address/v/1`	$$34.4 \mathrm{ms} \pm 258 \mathrm{μs}\left({\color{gray}4.07 \mathrm{\%}}\right) $$	Flame Graph

representative_read_entity_type

Function	Value	Mean	Flame graphs
get_entity_type_by_id	Account ID: `bf5a9ef5-dc3b-43cf-a291-6210c0321eba`	$$8.44 \mathrm{ms} \pm 36.8 \mathrm{μs}\left({\color{gray}-0.052 \mathrm{\%}}\right) $$	Flame Graph

representative_read_multiple_entities

Function	Value	Mean	Flame graphs
entity_by_property	traversal_paths=0	0	$$89.9 \mathrm{ms} \pm 440 \mathrm{μs}\left({\color{gray}0.707 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=255	1,resolve_depths=inherit:1;values:255;properties:255;links:127;link_dests:126;type:true	$$144 \mathrm{ms} \pm 525 \mathrm{μs}\left({\color{gray}0.522 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:0;link_dests:0;type:false	$$96.3 \mathrm{ms} \pm 389 \mathrm{μs}\left({\color{gray}-0.095 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:1;link_dests:0;type:true	$$107 \mathrm{ms} \pm 402 \mathrm{μs}\left({\color{gray}0.158 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:2;links:1;link_dests:0;type:true	$$114 \mathrm{ms} \pm 434 \mathrm{μs}\left({\color{gray}-0.003 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:2;properties:2;links:1;link_dests:0;type:true	$$123 \mathrm{ms} \pm 494 \mathrm{μs}\left({\color{gray}-0.475 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=0	0	$$101 \mathrm{ms} \pm 315 \mathrm{μs}\left({\color{gray}-0.236 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=255	1,resolve_depths=inherit:1;values:255;properties:255;links:127;link_dests:126;type:true	$$133 \mathrm{ms} \pm 414 \mathrm{μs}\left({\color{gray}1.90 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:0;link_dests:0;type:false	$$108 \mathrm{ms} \pm 386 \mathrm{μs}\left({\color{gray}0.501 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:1;link_dests:0;type:true	$$119 \mathrm{ms} \pm 447 \mathrm{μs}\left({\color{gray}0.338 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:2;links:1;link_dests:0;type:true	$$121 \mathrm{ms} \pm 545 \mathrm{μs}\left({\color{gray}1.65 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:2;properties:2;links:1;link_dests:0;type:true	$$120 \mathrm{ms} \pm 469 \mathrm{μs}\left({\color{gray}1.10 \mathrm{\%}}\right) $$

scenarios

Function	Value	Mean	Flame graphs
full_test	query-limited	$$190 \mathrm{ms} \pm 596 \mathrm{μs}\left({\color{gray}-0.608 \mathrm{\%}}\right) $$	Flame Graph
full_test	query-unlimited	$$173 \mathrm{ms} \pm 2.57 \mathrm{ms}\left({\color{lightgreen}-5.285 \mathrm{\%}}\right) $$	Flame Graph
linked_queries	query-limited	$$40.8 \mathrm{ms} \pm 162 \mathrm{μs}\left({\color{gray}0.239 \mathrm{\%}}\right) $$	Flame Graph
linked_queries	query-unlimited	$$541 \mathrm{ms} \pm 1.04 \mathrm{ms}\left({\color{gray}-3.066 \mathrm{\%}}\right) $$	Flame Graph

fix: estimation for aggregates

ae2d3af

vercel Bot temporarily deployed to Preview – petrinaut May 4, 2026 14:23 Inactive

github-actions Bot added area/libs Relates to first-party libraries/crates/packages (area) type/eng > backend Owned by the @backend team area/tests New or updated tests labels May 4, 2026

graphite-app Bot assigned indietyp May 4, 2026

indietyp mentioned this pull request May 4, 2026

BE-512: HashQL: Allow cross-backend transitions for SwitchInt terminators #8696

Open

3 tasks

indietyp force-pushed the bm/be-513-hashql-size-estimation-aggregate-cardinality-is-wrong-for branch from d52594c to ae2d3af Compare May 4, 2026 14:24

indietyp force-pushed the bm/be-512-hashql-switchint-allow-cross-backend-transitions branch from fb7f8ba to 5eeec53 Compare May 4, 2026 14:24

vercel Bot temporarily deployed to Preview – petrinaut May 4, 2026 14:24 Inactive

indietyp mentioned this pull request May 4, 2026

BE-525: HashQL: Remove implicit type-widening from Integer operations, and make comparison size aware #8693

Open

3 tasks

augmentcode Bot reviewed May 4, 2026

View reviewed changes

Comment thread libs/@local/hashql/mir/src/pass/analysis/size_estimation/footprint.rs

cursor Bot reviewed May 4, 2026

View reviewed changes

Comment thread libs/@local/hashql/mir/src/pass/analysis/size_estimation/estimate.rs

vercel Bot deployed to Preview – hash May 4, 2026 14:31 View deployment

fix: suggestion from code review

cf85692

vercel Bot temporarily deployed to Preview – petrinaut May 4, 2026 14:42 Inactive

github-advanced-security AI found potential problems May 4, 2026

View reviewed changes

Comment thread libs/@local/hashql/mir/src/pass/analysis/size_estimation/footprint.rs Fixed

vercel Bot deployed to Preview – hash May 4, 2026 14:51 View deployment

fix: lint

175adcb

vercel Bot deployed to Preview – petrinaut May 4, 2026 14:55 View deployment

vercel Bot deployed to Preview – hash May 4, 2026 15:00 View deployment

chore: truncate instead of resize on mul

598d416

vercel Bot temporarily deployed to Preview – petrinaut May 4, 2026 15:25 Inactive

vercel Bot deployed to Preview – hash May 4, 2026 15:33 View deployment

cursor Bot reviewed May 4, 2026

View reviewed changes

Conversation

indietyp commented May 4, 2026

🌟 What is the purpose of this PR?

🔍 What does this change?

🛡 What tests cover this?

❓ How to test this?

Uh oh!

vercel Bot commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cursor Bot commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Summary

Uh oh!

indietyp commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

augmentcode Bot commented May 4, 2026

Uh oh!

augmentcode Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

codspeed-hq Bot commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Footnotes

Uh oh!

codecov Bot commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Uh oh!

cursor Bot left a comment

Choose a reason for hiding this comment

Uh oh!

cursor Bot May 4, 2026

Choose a reason for hiding this comment

Empty cardinality treated as unbounded in materialize

Uh oh!

github-actions Bot commented May 4, 2026

Benchmark results

@rust/hash-graph-benches – Integrations

policy_resolution_large

policy_resolution_medium

policy_resolution_none

policy_resolution_small

read_scaling_complete

read_scaling_linkless

representative_read_entity

representative_read_entity_type

representative_read_multiple_entities

scenarios

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

2 participants

vercel Bot commented May 4, 2026 •

edited

Loading

cursor Bot commented May 4, 2026 •

edited

Loading

indietyp commented May 4, 2026 •

edited

Loading

codspeed-hq Bot commented May 4, 2026 •

edited

Loading

codecov Bot commented May 4, 2026 •

edited

Loading