Make commit numbers in version_performance_control.md links to the commit they refer to

[Happypig375]

And time to add a new column - arbitrary precision of trig functions cost a lot in performance

[WhiteBlackGoose]

Make commit numbers in version_performance_control.md links to the commit they refer to

I don't think it's big issue to be honest.
If someone wants to close it, they're welcome.

[WhiteBlackGoose]

File of interest

[dasingh92]

@WhiteBlackGoose I couldn't find the commit numbers in the file of interest. Can I have a better explanation of the outcome from this issue ?

[Rafael-SOWNet]

Half of this is landed in #795 (f207e37f), and the other half wants a decision.

Done: the commit numbers are links — they already were, and the file also documents how to count one (git rev-list --count <commit>). And the table has a new column, the 1620th, its first since the 1446th: the Kernel Benchmark filtered on a path that does not exist and had not run since 2026-01-02, so ninety-eight merged PRs went unmeasured (#775, fixed in #789).

Left: "a new column - arbitrary precision of trig functions cost a lot in performance". That needs a benchmark that does not exist yet rather than a measurement of one that does — CommonFunctionsInterVersion has no trigonometric case at all, at any precision. Adding one is straightforward, but it starts with a single data point and no history, so it is worth deciding whether you want it as a row in this table or as its own thing.

Happy to add the benchmark if you want it here. Leaving the issue open for that call rather than closing on the half that is done.

[Happypig375]

Sure, add as many benchmarks as needed, maybe profile memory use as well