RunEnd row-encode kernel (vortex-runend)#8008
Closed
joseph-isaacs wants to merge 1 commit into
Closed
Conversation
Add a row-encode kernel for `RunEnd` arrays via the inventory-based registry: the encoding lives in `vortex-runend` which depends on `vortex-array` (not the other way around), so a direct downcast inside `dispatch_size` / `dispatch_encode` would create a cycle. The kernel is functionally analogous to the Dict kernel: encode each unique run-value once into a small per-value buffer, then broadcast the value's encoded bytes across each row in its run. The per-unique-value cost is amortized over the number of runs rather than the row count. `walk_runs` translates the run-end array's `(prev_end, curr_end)` windows into `(start_logical, stop_logical)` row ranges accounting for the array's slice offset and length. When ends.len() > len (very sparse runs, or pathological inputs) the kernel declines so canonicalization stays the dominant path. Includes a round-trip test in `compute/row_encode.rs` checking that the RunEnd path matches the canonical path bit-for-bit. Signed-off-by: Claude <noreply@anthropic.com>
This was referenced May 18, 2026
This was referenced May 18, 2026
Closed
Merging this PR will degrade performance by 14.14%
|
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ❌ | Simulation | new_alp_prim_test_between[f32, 16384] |
103.8 µs | 118.4 µs | -12.33% |
| ❌ | Simulation | new_alp_prim_test_between[f32, 32768] |
153.1 µs | 182.1 µs | -15.91% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing claude/row-c23-runend-kernel (d8b9bce) with claude/row-c22-patched-kernel (b86974e)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Part 23 of 25 in the stacked PR series adding
vortex-row.This PR contains exactly one commit; review just that diff in isolation.
What this commit does
Adds a row-encode kernel for
RunEndarrays via the inventory-based registry: the encoding lives invortex-runendwhich depends onvortex-array(not the other way around), so a direct downcast insidedispatch_size/dispatch_encodewould create a cycle.The kernel is functionally analogous to the Dict kernel: encode each unique run-value once into a small per-value buffer, then broadcast the value's encoded bytes across each row in its run. The per-unique-value cost is amortized over the number of runs rather than the row count.
walk_runstranslates the run-end array's(prev_end, curr_end)windows into(start_logical, stop_logical)row ranges accounting for the array's slice offset and length. Whenends.len() > len(very sparse runs, or pathological inputs) the kernel declines so canonicalization stays the dominant path.Includes a round-trip test checking that the RunEnd path matches the canonical path bit-for-bit.
Stack
claude/row-c01-crate-scaffoldingclaude/row-c02-sortfield-optionsclaude/row-c03-codec-fixed-widthclaude/row-c04-codec-varlenclaude/row-c05-codec-nestedclaude/row-c06-rowsize-scalarfnclaude/row-c07-rowencode-scalarfnclaude/row-c08-convert-columns-tests-benchclaude/row-c09-skip-listview-validationclaude/row-c10-validity-fast-pathclaude/row-c11-skip-zero-initclaude/row-c12-vectorize-pure-fixed-offsetsclaude/row-c13-vectorize-mixed-offsetsclaude/row-c14-varlen-block-copy-nonoverlappingclaude/row-c15-walk-varbinview-directlyclaude/row-c16-arith-write-fast-pathclaude/row-c17-specialize-constant-arithclaude/row-c18-kernel-dispatch-helpersclaude/row-c19-inventory-registryclaude/row-c20-constant-kernelclaude/row-c21-dict-kernelclaude/row-c22-patched-kernelclaude/row-c23-runend-kernelclaude/row-c24-bitpacked-kernelclaude/row-pr3-kernelsBase of this PR: #8007 (
claude/row-c22-patched-kernel)Next in stack: #8009 (
claude/row-c24-bitpacked-kernel)Combined context
For the full design + rationale, see PR #7985 (top of stack).