This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]

Re: Zero-length arrays

To: Linus Torvalds <torvalds at transmeta dot com>
Subject: Re: Zero-length arrays
From: Richard Henderson <rth at redhat dot com>
Date: Thu, 4 Jan 2001 10:54:38 -0800
Cc: dkorn at pixelpower dot com, gcc at gcc dot gnu dot org
References: <718D38CAB6E0D011B2C90060970C28A5642562@EXCHANGESERVER> <200101041837.KAA15911@penguin.transmeta.com>

On Thu, Jan 04, 2001 at 10:37:44AM -0800, Linus Torvalds wrote:
> (I never understood why C had that silly "you can't have a zero-length
> array" rule. Even if there were no real reasons to use them, it still
> has a clear and unambiguous conceptual meaning)

Mostly because they are petrified of objects with zero length, I think.

Not without reason -- they can cause much pain.  I havn't thought about
all the ways things can fail, but have run into a couple in playing
around.  None serious enough that I thought it absolutely necessary that
we restrict zero-length arrays to otherwise non-empty structures, but
nevertheless the stance gcc is going to have to take in such edge cases
is Don't Do That.

r~

References:
- RE: Zero-length arrays
  - From: David Korn
- Re: Zero-length arrays
  - From: Linus Torvalds

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]