bug#68244: hash-table improvements

[Stefan Monnier]

> But let's try anyway: here is a run of one of the main suites I've been
> using. The numbers indicate relative changes in percent of elapsed time from
> the baseline to the tip of the scratch/hash-table-perf branch, negative
> numbers being better. For example, -50 means that speed has doubled, +100
> that it has halved.
>
> [2. application/vnd.lotus-organizer; pct.org]

Lotus, eh?

> |                           | 681a2877cc2 |
> |---------------------------+-------------|
> | make empty                |       -88.9 |
> | make 4, dead              |       -88.8 |
> | make 16, dead             |       -62.3 |
> | make 64, dead             |       +25.1 |
> | make 256, dead            |       -77.7 |
> | make 1024, dead           |       -76.7 |
> | make 4096, dead           |       -81.4 |
> | make 100000, dead         |       -73.2 |
> |---------------------------+-------------|
> | make 4, alive             |       -93.3 |
> | make 16, alive            |       -75.6 |
> | make 64, alive            |       +11.5 |
> | make 256, alive           |       -82.4 |
> | make 1024, alive          |       -79.7 |
> | make 2048, alive          |       -83.3 |
> | make 4096, alive          |       -84.4 |
> | make 100000, alive        |       -75.1 |

What's up with 64?


        Stefan