[plt-scheme] nonstandard installation

From: Eli Barzilay (eli at barzilay.org)
Date: Sat Nov 29 00:52:57 EST 2008

On Nov 28, Prabhakar Ragde wrote:
> Matthew Flatt wrote:
> > Are these Sparc machines or x86 machines? Sparc would explain
> > slowness on Robby's test, because the JIT doesn't currently
> > support Sparc --- and Robby's test is a worst-case example for JIT
> > versus non-JIT execution.
> These are Sparc machines. There are some x86 Linux servers in the
> undergraduate environment; I will request PLT Scheme installation on
> those.
> > For start-up, my guess is that the difference is in where the
> > bytecode files are stored, since start-up time is dominated by
> > loading in bytecode. Another possibility (which also applies to
> > the difference in the scaled-down version of Robby's test) is a
> > difference between using shared libraries versus static builds, if
> > one the other version was configured with --enable-shared.
> I will look into the location of the .zo files (and where the user
> directory containing the locally-built version is stored), and ask
> about the installation configuration options. Thanks. --PR

JFYI, my experience with sparc-solaris machines in a few places is
that using NFS can make things extremely slow (and if you see high
system time, then that's most likely it).  You can try to copy it to a
machine local directory (like /var/tmp -- I think that on most solaris
configurations /tmp is a memory FS) to factor out the NFS delay.

          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                  http://www.barzilay.org/                 Maze is Life!

Posted on the users mailing list.