Re: [Gcl-devel] Version_2_7_0t4

[Camm Maguire]

Greetings!  Thanks as always Mike for your incredible efforts here.  I
really apologize if recent developments have made your life more
difficult.  I think the immediate fixnum and two word cons which they
support ar eworth the effort on mingw -- certainly (hopefully) Jared
at the University of Texas woudl concur!

"Mike Thomas" <address@hidden> writes:

> Hi Camm/Vadim.
> 
> | Greetings!  Happy to announce:
> ...
> | Testing version: (cvs tag) Version_2_7_0t4
> 
> Congratulations Camm.
> 
> I thought I should bring you up to date on wher I am at.
> 
> I've been getting into GCL again over the past week and I'm having great
> difficulty with HEAD on Windows.  I have been unable to build HEAD since
> about 11 June, due to the memory configuration/garbage collection/allocation
> changes at that time.
> 
> The complexity of the system has gotten out of hand from my point of view
> and I am still groping at trying to understand the meaning of most of the
> configuration parameters and/or how they are derived in the configure
> script.

I likely need to summarize this to the list -- when I get a moment.
The basic idea is to get as much heap space as possible after
determining any limits set by the stack location and shared memory
addresses, then set an immediate fixnum table in the inaccessible range.

Take care,

> 
> Because of limited time I will focus on this problem until it is fixed, but
> I would particularly appreciate some help from you, Vadim, on this problem
> as you seem to understand the memory management of GCL better than I, and
> also build on Windows sometimes.
> 
> Cheers
> 
> Mike Thomas.
> 
> 
> 
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah