jerrinot's improvement - everyone is hunting ILP? let's do less!

[jerrinot]

it turns out doing 3 things at once was too much. perf annotate showed spilling.

Check List:

• You have run ./mvnw verify and the project builds successfully
• Tests pass (./test.sh <username> shows no differences between expected and actual outputs)
• All formatting changes by the build are committed
• Your launch script is named calculate_average_<username>.sh (make sure to match casing of your GH user name) and is executable
• Output matches that of calculate_average_baseline.sh
• For new entries, or after substantial changes: When implementing custom hash structures, please point to where you deal with hash collisions (line number)

• Execution time: 1.8s
• Execution time of reference implementation: long

thanks to @franz1981 for the idea to explore register spilling and to @thomaswue for tips on how to do it!

[gunnarmorling]

@jerrinot, where does it handle hash collisions actually?

[jerrinot]

closing in favour of #690

[jerrinot]

@gunnarmorling: it uses 2 maps:

1. slow map, which can store long names. the name is checked here: https://github.com/jerrinot/1brc/blob/f76e8d27c706cc785a79c6f598202a0d9d5e2566/src/main/java/dev/morling/onebrc/CalculateAverage_jerrinot.java#L571-L575
2. fast-map. which can store only names up to 15 bytes. and it's returning only when both words are matching: https://github.com/jerrinot/1brc/blob/f76e8d27c706cc785a79c6f598202a0d9d5e2566/src/main/java/dev/morling/onebrc/CalculateAverage_jerrinot.java#L519-L521

[gunnarmorling]

Got it, thx!

…

Message ID: ***@***.***>