LuciferYang opened a new pull request, #13144: URL: https://github.com/apache/gluten/pull/13144
## What changes are proposed in this pull request? `ReadTimeAccumulator` converted each per-call read duration from nanoseconds to milliseconds before accumulating (`onAdded(TimeUnit.NANOSECONDS.toMillis(after - prev))`). For batched columnar iterators each `hasNext`/`next` runs well under a millisecond, so every call truncated to 0 and the collected read-time metric stayed near zero regardless of the real work done. This accumulates nanoseconds and reports only whole milliseconds, carrying the sub-millisecond remainder over to the next call, so the metric reflects the actual total read time. ## How was this patch tested? New `IteratorSuite` case iterates 300k elements through the wrapped iterator and asserts the accumulated read time is greater than zero; the previous per-call truncation would have reported zero. ## Was this patch authored or co-authored using generative AI tooling? Yes. Generated-by: Claude Opus. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
