zhiqiang-hhhh commented on PR #64377: URL: https://github.com/apache/doris/pull/64377#issuecomment-4890917829
Overall the PR looks good to me. A couple of detail points worth discussing: 1. The predicate-column reads that are eagerly executed here could potentially be skipped by the subsequent main scan loop. As it stands, after `_eager_filter_predicates_into_bitmap` narrows `_row_bitmap`, the predicates are neither cleared nor marked resolved, so the main scan loop re-reads the same predicate columns and re-evaluates the predicates over the narrowed set — i.e. the predicate columns are read/evaluated twice. Might be worth exploring whether the main loop can reuse this result and skip the redundant read/eval. 2. It might be nice to add a counter in the AnnIndex section of the profile for the number of rows filtered out by this pre-filter pass. Right now this eager pass doesn't update any stats (e.g. `blocks_load` / `raw_rows_read`), so the work it does is invisible in the profile. -- 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]
