From owner-cvs-all  Mon Feb  8 09:33:50 1999
Return-Path: <owner-cvs-all@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id JAA02194
          for cvs-all-outgoing; Mon, 8 Feb 1999 09:33:50 -0800 (PST)
          (envelope-from owner-cvs-all@FreeBSD.ORG)
Received: from lor.watermarkgroup.com (lor.watermarkgroup.com [207.202.73.33])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA02177;
          Mon, 8 Feb 1999 09:33:38 -0800 (PST)
          (envelope-from luoqi@watermarkgroup.com)
Received: (from luoqi@localhost)
	by lor.watermarkgroup.com (8.8.8/8.8.8) id MAA12554;
	Mon, 8 Feb 1999 12:33:28 -0500 (EST)
	(envelope-from luoqi)
Date: Mon, 8 Feb 1999 12:33:28 -0500 (EST)
From: Luoqi Chen <luoqi@watermarkgroup.com>
Message-Id: <199902081733.MAA12554@lor.watermarkgroup.com>
To: luoqi@watermarkgroup.com, newton@atdot.dotat.org
Subject: Re: cvs commit: src/sys/kern imgact_elf.c
Cc: cvs-all@FreeBSD.ORG, cvs-committers@FreeBSD.ORG, newton@FreeBSD.ORG
Sender: owner-cvs-all@FreeBSD.ORG
Precedence: bulk

> Luoqi Chen wrote:
> 
>  > None of the executables under usr/ on the CD-ROM are statically linked, so
>  > you don't need to brand them at all. The problem is that the interpreter
>  > name in the svr4 elf brand structure is *WRONG*, it should be
>  > /lib/ld.so.1, not /lib/libc.so.1 (svr4/svr4_sysvec.c).
> 
> This is not correct.  Firstly, SysV considers that libc is the "interpreter"
> (dump out some SysV ELF objects to check if you like).  Secondly,
> making the change as you suggest simply doesn't work; executables 
> still require branding.
> 
> 
>     - mark
> 
It seems that not all SysV systems use the same ELF interpreter, each one
of us sees a different one :( There is a solution to this situation though,
register multiple brandinfo structures, one for each possible interpreter,
just like how linux emulator handles libc5 and glibc.

-lq

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message