Re: no constants please

[Michael Sperber <sperber@xxxxxxxxxxxxxxxxxxxxxxxxxxx>]

>>>>> "Tom" == Tom Lord <lord@xxxxxxx> writes:

>> From: Michael Sperber <sperber@xxxxxxxxxxxxxxxxxxxxxxxxxxx>

Tom> If the root set is large, certainly it should be traced in several
Tom> steps, using barriers to preserve its invariants.

>> Is there a practical example of a system that does this?  It seems
>> very difficult to do, even absent an FFI to C, as your typical root
>> set---the current continuation---changes *all the time*.  (I'm really
>> curious.  I could never wrap my mind around this.)

Tom> You can treat the "big-three abstract registers" (continuation, code,
Tom> and environment) specially.   They have usefully limited usage
Tom> patterns.   It's the other roots, if your implementation has them,
Tom> that are of greater interest.  (The draft FFI creates "other roots".)

That isn't the question I asked.  All hard questions are buried behind
"specially."

Tom> If you want to have a chat about incremental GC strategies, please
Tom> c'mon over to the pika-dev@xxxxxxxxxxx mailing list:

It seems the GC list would be the right place to discuss this.

Tom> Not all incremental collectors are incompatible with the FFI (a mostly
Tom> copying semi-conservative incremental GC would be one example).

This was actually a precise non-copying incremental GC.

-- 
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla