Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!charles
From: charles@sdd.hp.com (Charles Keith)
Newsgroups: comp.os.minix
Subject: Re: 1.3->1.5, readclock fails on AT.
Keywords: dev/port, readclock, ls
Message-ID: <1990May3.004703.1113@sdd.hp.com>
Date: 3 May 90 00:47:03 GMT
References: <365@cancol.oz>
Organization: Hewlett Packard, San Diego Division
Lines: 36


In article <365@cancol.oz> eyal@cancol.oz (Eyal Lebedinsky) writes:
>
>ls had a minor error (_21 multiple ets, normal asld problem) so I removed
>the label. For this I libunpacked ls.s then libpacked back. Libpack said the
>...
>
>The questions here are: is there a problem with doing unpack/pack in parts?
>I thought the packing was a one to one string replacement. Also, is there a
>solution to the asld problem like this one?
>
Exactly the same problem happened to me.  Rather than trying to re-pack ls.s,
I just fed the unpacked file to asld (this is allowed).  I got rid of the
multiple define, but the resulting binary dumped core doing ls -l (other
options ran ok).  The rogue label _21 seemed to be in the vicinity of the
static array char *rwx (also associated with the -l option. hmmm..).
I moved it out of the function it was in, and ls compiled and linked and ran ok.

  Since this problem hasn't been widely reported, I assume it must be an
artifact of the 1.2 compiler, since I assume all the gurus are using the 1.3
compiler.

 In doing all of this, I spent a lot of time looking at the symbol tables for
ls, and I was able to learn very little about what gets linked into the
executable.  Most compiler/linkers will give a pretty good map when asked,
minix seems weak in this area.  Am I missing something here?  Before I fix a
problem I like to UNDERSTAND it, and in this case I was reduced to trial and
error.  Any helpful hints or tutorials out there on DEBUGGING under minix?
How can problems like this be handled scientifically?  Or do we just sprinkle
liberally with printf's, bake and test?

-- 
Charles Keith
Hewlett-Packard Co.         UUCP: {hplabs|ucsd|hpfcla}!hp-sdd!charles
16399 W. Bernardo Dr.       Internet: charles%hp-sdd@hplabs.HP.COM
San Diego, CA 92127         Disclaimer: I am not a qualified spokesperson.