*BSD News Article 6789


Return to BSD News archive

Xref: sserve comp.unix.bsd:6838 comp.lang.perl:12052
Newsgroups: comp.unix.bsd,comp.lang.perl
Path: sserve!manuel.anu.edu.au!munnari.oz.au!uunet!snorkelwacker.mit.edu!bloom-picayune.mit.edu!athena.mit.edu!eichin
From: eichin@athena.mit.edu (Mark W. Eichin)
Subject: Re: Installing perl-4.035 in 386BSD
In-Reply-To: adrian@rachel.ibmpcug.co.uk's message of 20 Oct 92 08:59:12 GMT
Message-ID: <EICHIN.92Oct20133705@tsx-11.mit.edu>
Sender: news@athena.mit.edu (News system)
Nntp-Posting-Host: tsx-11.mit.edu
Organization: Massachusetts Institute of Technology
References: <IKEDA.92Oct19202131@phenix.ntec.co.jp> <BwEwyp.DF1@ibmpcug.co.uk>
Date: Tue, 20 Oct 1992 17:37:18 GMT
Lines: 17

Gee, I had no problem compiling perl-4.035 with -O; I had 16M of
physical memory, and no swap (well, no working swap). Important trick:
from the shell, "unlimit datasize" "unlimit stacksize" etc.
	Note, however, that perl will not pass all of its tests (I
posted about this once before.) Most of them are simple differences in
the system (in particular, the dbm ones) however the bigint tests fail
badly... from some email I've received it seems that the reason is
that there are bugs in the 386BSD floating point emulation. 
	I've hacked up set of soft-float routines for gcc (they'll
be contributed to the FSF some time soon, and probably get cleaned up
a lot) that are good enough to pass Kahan's "paranoia" test (this
requires getting rounding *exactly* right and things like that.) I'd
been meaning to try building perl (with these routines, perhaps using
a cross-compiler) but I'm off hacking networking in Linux for now :-)
				_Mark_ <eichin@athena.mit.edu>
				MIT Student Information Processing Board
				Cygnus Support <eichin@cygnus.com>