emecii commented on PR #11052: URL: https://github.com/apache/arrow-rs/pull/11052#issuecomment-5690699961
For #11052, I can add a benchmark-only PR covering list-index validity: in-bounds/missing indexes, explicit nulls, nested lists, List/ListView, sliced inputs, typed/Variant outputs, and a struct-only control. I'd compare identical fixtures on base/head and measure allocations separately from runtime. An initial local allocation probe shows additional bitmap-union costs on nullable paths, so shared-buffer cloning alone doesn't establish the cost. Does this cover the evidence you'd like before merging the fix? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
