Path: utzoo!attcan!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!samsung!sol.ctr.columbia.edu!emory!mephisto!ukma!usenet.ins.cwru.edu!eagle!news
From: mckim@mildred.lerc.nasa.gov (Jim McKim)
Newsgroups: comp.unix.sysv386
Subject: Re: Which 386 UNIX binaries run on which 386 UNIXs ?
Message-ID: <1990Sep14.191716.29324@eagle.lerc.nasa.gov>
Date: 14 Sep 90 19:17:16 GMT
References: <1990Sep5.171859.21535@eagle.lerc.nasa.gov> <1990Sep6.080554.20815@fiver>
Reply-To: mckim@mildred.lerc.nasa.gov (Jim McKim)
Organization: NASA/Lewis Research Center, Cleveland
Lines: 27

In article <1990Sep6.080554.20815@fiver> palowoda@fiver (Bob Palowoda) writes:
>  Can you be a little more specific about what the binaries are? Maybe
>'what' libs?  Actually I was quite suprised to find out alot of my ESIX
>binaries ran under Xenix 386. I poked around and couldn't find the 
>"bad system call" error message. Where does this come from?

I copied the g++.tar.Z file from sequent.kent.edu, installed the stuff in
this file, then tried to compile something.  The preprocessor
/usr/local/lib/gcc-cpp failed on my intel/unix machine.

The error message (from the shell) was "Bad system call - core dumped".

The file (gcc-cpp) is stripped.  Postmortem sdb reports the machine was
executing (about 6 frames deep at addr 0x374f)
	"movb   $0x00,(%ebx,%eax,1)      [(%ebx,%eax,1)]"
%eax was 0 and %ebx was 0x406048.

This looks like moving a byte from here to there, not a system call.
System calls come out as "lcall"s in a disassembly.  I am probably
misinterpreting something.

This binary only uses my shared c library (libc_s).
--
----------------
Jim McKim  /  Internet: mckim@mildred.lerc.nasa.gov		"" -
Phone: +1 216 891 2977  /  Packet: kb8dcr@kb8dcr.ampr.org	Needermeyer
----------------