Re: lazy sweeping.

[Mikael Djurfeldt]

Han-Wen wrote:

> Here are some results of my lazy sweeping implementation:
>
> [...]
>
> The lazy sweeping approximately halves the startup time of GUILE. I
> consider the old GC beat; when can we start integrating this? 
Nice! However, be careful about benchmarking information. When hacking
the evaluator, I've noticed performance changes up to 30% just by changing
the textual order of functions within a compilation unit...

> * Lazy sweeping naturally complements multi-threaded use: whenever a
>   thread runs out of free cells, we can simply sweep a few (say 512)
>   cells, and return that for private use by that thread.
[Without having studied your code:] Is it enough to lock one mutex for the
joint apparatus of sweeping and marking before doing the lazy sweep
in order to avoid interference between threads?  (One would not want to
stop all threads before sweeping 512 cells...)

Best,
M