Hi,

I've just checked in (as sbcl-1.0.14.33) an update to a couple of
alpha runtime files, with the result that the system builds again on
linux; I've tested on 2.6.18 as in Debian etch.  I haven't done
extensive fixing, and in particular I suspect that sb-posix is broken,
but the general system should be in better health than it has been for
a year.  (There were about 4 different brokennesses, all fairly
trivial once I worked out what they actually were.)  Thanks to Paul
Khuong and the University of Alberta's maths department for making a
machine available to me.

I hope that this makes it possible for Debian to reinstate support for
sbcl/alpha in its buildd network: I view that as an invaluable
resource to help us avoid bitrot.  On a related subject, Thiemo has
mentioned that there may be trouble with SBCL on SPARC 2.6 kernels; I
only have access to a SPARC with 2.4, so I can't do any debugging
there (and I don't know what the issue is in any case).  Can anyone
clarify that?

This now puts me in a position to finish my work on the updated
modular arithmetic stuff, for choosing an appropriate implementation
where more than one is applicable.  I do not intend to get it in for
this month, but hopefully early next month that can land on HEAD and I
can forget about it...

Best,

Christophe

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/