This is the mail archive of the guile@sources.redhat.com mailing list for the Guile project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: GUILE's GC - why we struggling to solve already solved problems ? (was Re: finding the top of the stack in linux)


"Khimenko Victor" <guile@khim.sch57.msk.ru> writes:

> But why the hell we are inventing our own GC with a lot of
> limitations and not even using clearly appropriate parts (like stack
> placement finding subroitine) from ALREADY existing and TESTED
> solutions.

Guile's GC is just as much `there' as is Boehm's GC.  We are not
inventing it.

Furthermore, it is unclear to me whether Boehm's GC would be an
improvement over Guile's current GC.  While Guile's GC is technically
a conservative GC because it scans certain memory areas for possible
pointers, it is in fact mostly precise.  Do you think the Boehm GC can
be as efficient even when it is completely conservative?  The most
convincing way would be to produce a version of Guile with uses
Boehm's collector.

In general, my view is that the GC of Guile is in fact an area where
significant improvements can be made, but I see these improvements
mostly in the direction of using more `modern' variants of GC than
mark/sweep, like generational collectors, and mostly copying ones.  I
would regard switching to a purely conservative GC as a step in the
wrong direction.

On the other hand, Boehm knows what he is doing when it comes to GC,
I'd guess.  So maybe his collector really is superior.  I don't know.

In any case, we could probably raid it for useful portability code,
like code for finding the stack boundaries, if it contains something
like that.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]