[5516] in Athena Bugs
Re: Crummy error reporting from ld. (Athena RT release 7.0 with hc2.1s)
daemon@ATHENA.MIT.EDU (Bill Cattey)
Thu Jul 12 01:48:14 1990
Date: Thu, 12 Jul 90 01:47:56 -0400 (EDT)
From: Bill Cattey <wdc@ATHENA.MIT.EDU>
To: "Jonathan I. Kamens" <jik@pit-manager.MIT.EDU>
Cc: bugs@ATHENA.MIT.EDU
In-Reply-To: <9007111800.AA13194@pit-manager.MIT.EDU>
RE: 1. Please don't send bugs to ibmsupt directly.
Eric Brunner is a personal friend of mine and has specifically asked for
hc bug reports. I sent to Eric, and later decided to send it to bugs.
As I explained in person, there was no Z0 magic running because ld was
being called by a shell script written at CMU with no knowledge of Z0
magic.
The errors you saw were the entirety of the errors given. On the VAX,
only 2 multiply defined errors were printed. I was told by either jfc
or raeburn that this is because on the RT, a procedure call is really
two symbols to define.
I was able to find my problem using nm, sort, and uniq. It turned out
that some bad coder was defining procedures in a .h file that I included
in two .c files that made two .o files which got linked into my dynamic
executable. There really were two procedures multiply defined.
-wdc
