Re: [MIT-Scheme-devel] Symmetric MultiProcessing

[Chris Hanson]

R7RS defines parameters, which we'll eventually need to implement anyway for compliance.  Converting the system's bindings over to parameters is probably the right thing for a variety of reasons.

On Fri, Jan 17, 2014 at 5:49 AM, Taylor R Campbell <address@hidden> wrote:

   Date: Thu, 16 Jan 2014 20:00:54 -0700
   From: Matt Birkholz <address@hidden>

   If we want SMP(?) and don't want it in a distant fork, we might just
   bite the bullet and replace our fluid bindings with fluid objects
   (like e.g. s48's).  That's an assq on each reference but the thread-
   local values are stored and accessed in a straightforward manner, and
   there is no winding and unwinding in state-space:global at every
   thread switch.

That's what I have been intending to do for ages.  For large sets of
fluids like the compiler uses, we could merge them into one fluid with
a large data structure.

We might use a balanced binary tree instead of an alist, but it
probably won't make a big difference.  Also, it might be worth looking
into what Racket calls `parametrizations'.

_______________________________________________
MIT-Scheme-devel mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/mit-scheme-devel