From mboxrd@z Thu Jan  1 00:00:00 1970
From: Mark Mitchell <mark@codesourcery.com>
To: ian@zembu.com
Cc: rth@cygnus.com, binutils@sourceware.cygnus.com
Subject: Re: Patches for IRIX6 N32-ABI ld
Date: Thu, 01 Jul 1999 00:00:00 -0000
Message-id: <19990626134622L.mitchell@codesourcery.com>
References: <19990626112510.A21222@cygnus.com> <19990626120519O.mitchell@codesourcery.com> <19990626192703.9669.qmail@daffy.airs.com> <19990626192703.9669.qmail@daffy.airs.com>
X-SW-Source: 1999-q2/msg00358.html

>>>>> "Ian" == Ian Lance Taylor <ian@zembu.com> writes:

    Ian> I for one would be much more comfortable with testing a
    Ian> boolean expression rather than writing a test based on the
    Ian> assumption that an enum value is and will remain zero.  It
    Ian> seems too easy for somebody to change the enum and thus
    Ian> unexpectedly change the behaviour of the code.  I believe the
    Ian> patch as written introduces a maintenance pitfall.

Wow, I'd hoped only to get controversial at the time we started
talking about relocation processing and stuff like that.  :-) 

    Ian> In other words, I think either SGI_COMPAT should continue to
    Ian> be defined as a clearly boolean value (e.g., 1) or you should
    Ian> modify all the tests of SGI_COMPAT to work in some different
    Ian> manner.

I understand what you're saying.  But, this is a common programming
idiom, and well understood.  How about explicitly setting `sct_none'
to zero:

  enum {
    sct_none = 0,
    ...
  }

together with a comment?

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com