[Openmcl-devel] Building from CVS

David Steuber david at david-steuber.com
Tue Apr 5 20:31:20 PDT 2005

On Apr 5, 2005, at 8:51 PM, Gary Byers wrote:

> There have been some pretty drastic ABI changes in the bleeding-edge
> tree, and you do need a kernel compiled from CVS + the most recent
> image from the /testing directory to compile current bleeding-edge
> sources.

I went to fetch a dppcl.image from the ftp server on clozure.com, but 
all I could find was this:

-rw-r--r--    1 1001     0         1936124 Mar 28 05:13 

That was the newest file and seems to be for Linux.  It doesn't work on 
the kernel compiled for Darwin.  I didn't see a corresponding Darwin 
version apart from the old ones.

Out of curiosity, how do you deal with these chicken and egg situations 
during development?  The need to fetch a binary to build OpenMCL from 
CVS seems to be a weak link.  Of course my current build process isn't 
very robust either as I compile the kernel first and then attempt to 
use an old image on it to build a new image.  I should probably fix 
that except the old kernel with old image wasn't able to build a new 
image because of that unrecognized opcode.

More information about the Openmcl-devel mailing list