[GH-ISSUE #3240] LowerName and LowerQuery may be redundant #1158

New issue

Closed

opened 2026-03-16 01:45:35 +03:00 by kerem · 2 comments

kerem commented

2026-03-16 01:45:35 +03:00

Owner

Originally created by @divergentdave on GitHub (Aug 28, 2025).
Original GitHub issue: https://github.com/hickory-dns/hickory-dns/issues/3240

LowerName is similar to Name, and is in fact a newtype around Name, with its new() constructor converting names to lowercase. At this point, <Name as Ord> uses case-insensitive comparison, so converting to lowercase only has superficial impacts when displaying the name, etc. Similarly, LowerQuery wraps a LowerName and a Query, with the invariant that the name field is the lowercase form of the query's name. There is a comment that "LowerQuery is intended to reduce complexity for lookups in authorities", but I think that this is no longer the case, since all Name comparisons are case-insensitive. We could likely remove both wrappers, saving us from doing some conversions and eliminating some extra copies.

Originally created by @divergentdave on GitHub (Aug 28, 2025). Original GitHub issue: https://github.com/hickory-dns/hickory-dns/issues/3240 `LowerName` is similar to `Name`, and is in fact a newtype around `Name`, with its `new()` constructor converting names to lowercase. At this point, `<Name as Ord>` uses case-insensitive comparison, so converting to lowercase only has superficial impacts when displaying the name, etc. Similarly, `LowerQuery` wraps a `LowerName` and a `Query`, with the invariant that the name field is the lowercase form of the query's name. There is a comment that "LowerQuery is intended to reduce complexity for lookups in authorities", but I think that this is no longer the case, since all `Name` comparisons are case-insensitive. We could likely remove both wrappers, saving us from doing some conversions and eliminating some extra copies.

kerem closed this issue

2026-03-16 01:45:41 +03:00

kerem commented

2026-03-16 01:45:46 +03:00

Author

Owner

@bluejekyll commented on GitHub (Sep 3, 2025):

The reason I added LowerName was for performance. Given all the Name comparisons in the code, use a case-sensitive comparison vs. a case-insensitive comparison has a noticeable impact to micro-benchmark performance, and overall I think accounted for something on the order of a 10% boost in request latency.

@bluejekyll commented on GitHub (Sep 3, 2025): The reason I added LowerName was for performance. Given all the Name comparisons in the code, use a case-sensitive comparison vs. a case-insensitive comparison has a noticeable impact to micro-benchmark performance, and overall I think accounted for something on the order of a 10% boost in request latency.

kerem commented

2026-03-16 01:45:52 +03:00

Author

Owner

@divergentdave commented on GitHub (Sep 4, 2025):

Ah, that makes sense. I ran the benchmarks in the proto crate, and I get similar results.

@divergentdave commented on GitHub (Sep 4, 2025): Ah, that makes sense. I ran the benchmarks in the proto crate, and I get similar results.

kerem referenced this issue

2026-03-16 04:42:00 +03:00

[PR #1158] [MERGED] Support mDNS unicast-response bit #2023

kerem referenced this issue

2026-03-16 04:57:32 +03:00

[PR #1158] [MERGED] Support mDNS unicast-response bit #2028