Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!sri-spam!ames!ucbcad!ucbvax!SEAS.UCLA.EDU!bilbo.geoff
From: bilbo.geoff@SEAS.UCLA.EDU.UUCP
Newsgroups: comp.windows.x
Subject: Re: Help needed on SYSV Compiler Problem
Message-ID: <8711191225.AA27130@ATHENA.MIT.EDU>
Date: Wed, 18-Nov-87 22:08:43 EST
Article-I.D.: ATHENA.8711191225.AA27130
Posted: Wed Nov 18 22:08:43 1987
Date-Received: Sat, 21-Nov-87 15:43:06 EST
References: <JOEH@BNR.BITNET> <8711171555.AA29388@ATHENA.MIT.EDU>
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
Lines: 24

> We are porting X V11 to a SysV lookalike.  We actually have
> the SysV.1 and SysV.2 compilers, both of which fail with the same
> internal compiler error:
...
> Has anyone else encountered this or a similar problem?
> Any advice would be much appreciated.

We've gotten a couple of problems on SysV (Microport) here at Locus.
One is the old "too much defining" from the preprocessor;  the other
only arises when you compile with -g, which gives you "fakename table
overflow".  (The fakename table has all of 99 entries by default).

Both of these problems can be cured easily if you have source:
just grep for the offending message;  the corresponding limit will
be easy to spot in the nearby code.  You can then grep everything
to find where this limit is defined, and increase it.

If you don't have source, you can start by complaining to your vendor.
Since this rarely bears quick fruit, a (tedious) workaround is to
ifdef parts of header files, so that each routine only gets what it
needs.  You may have to break some routines like dispatch.c in order
to get the required part of the header files down far enough.

	Geoff Kuenning	geoff@lcc.ucla.edu	geoff@ITcorp.com