[comp.text.tex] Experience with Common TeX 3.0 on 386 UNIX

glenn@suphys.physics.su.OZ.AU (Glenn Geers) (11/27/90)

Hi,
	well I fetched Pat's new offering and compiled it on our SUN 4 and all went
well - produced TeX and LaTeX - and both work fine. I then had a go on our 386
box (Sys V.3.2) and the compile went smoothly. Make tex worked fine, but make
latex failed (Frank Mittelbach's lplain.tex same as on the Sun). Ok, probably
a problem with 'cc' producing something odd. Stranger things have happened.
So I recompile with 'gcc' (COFF/stabs) only to have the same thing happen 
again. 

Well anyway initex lplain dies just before it looks at hyphen.tex. A poke around
with the debugger shows that it dies in new_node() (alloc.c) which is called
from new_sym() (sym.c) as (when it crashes) new_sym("stackrel", 9) which seems
ok. Has anyone got any further? or is it worth my while to have a longer look?

					Cheers,
						Glenn

BTW Keep up the good work Pat.


glenn@qed.physics.su.oz.au


--
Glenn Geers                       | "So when it's over, we're back to people.
Department of Theoretical Physics |  Just to prove that human touch can have
The University of Sydney          |  no equal."
Sydney NSW 2006 Australia         |  - Basia Trzetrzelewska, 'Prime Time TV'

pcg@cs.aber.ac.uk (Piercarlo Grandi) (11/30/90)

On 26 Nov 90 22:31:40 GMT, glenn@suphys.physics.su.OZ.AU (Glenn Geers) said:

glenn> [ ... CTeX 3.0 goes fine on SPARC ... ]
glenn> [ ... but fials on LaTeX on 386 ... ]

Well, I have compiled the whole thing under ESIX, with 'cc', using
unexec() from GNU Emacs instead of undump, and it works for me. I used
the 3.0 LaTeX and TeX macros as found on labrea.stanford.edu.

glenn> Well anyway initex lplain dies just before it looks at
glenn> hyphen.tex. A poke around with the debugger shows that it dies in
glenn> new_node() (alloc.c) which is called from new_sym() (sym.c) as
glenn> (when it crashes) new_sym("stackrel", 9) which seems ok. Has
glenn> anyone got any further?

I had no problem whatsoever. Try using unexec() and not undump -- it is
more portable anyhow. I am even tuning it -- first memory allocation,
then the rest. I have sent back some patches to Pat Monardo, along with
the source of my superefficient malloc() clone. They should appear
eventually.
--
Piercarlo Grandi                   | ARPA: pcg%uk.ac.aber.cs@nsfnet-relay.ac.uk
Dept of CS, UCW Aberystwyth        | UUCP: ...!mcsun!ukc!aber-cs!pcg
Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.uk