Path: utzoo!utgpu!utstat!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!lll-winken!uunet!mcvax!ukc!acorn!ixi!clive
From: clive@ixi.UUCP (Clive)
Newsgroups: comp.lang.c
Subject: Re: Structure padding
Summary: Nice try, but wrong
Message-ID: <137@ixi.UUCP>
Date: 13 Apr 89 06:31:01 GMT
References: <440@front.se> <8392@xanth.cs.odu.edu> <124@paix.ikp.liu.se>
Reply-To: clive@ixi.uucp (Clive Feather)
Organization: none
Lines: 52
Expires:
Sender:
Followup-To:

In article <124@paix.ikp.liu.se> pekka@paix.ikp.liu.se (Pekka Akselin [The Mad Midnight Hacker]) writes:
>>>[...]
>>>I need to know if I can safely assume that there will not be any
>>>padding between the arrays.
>
>How about this???
>
>union {
>	struct foo {
>	    char	Bar[3];
>	    char	Baz[6];
>	    char	Frotz[4];
>	} fred;
>	char	SomeString[3 + 6 + 4];
>} Strings;
>
>#define bar	Strings.Bar
>#define baz	Strings.Baz
>#define frotz	Strings.Frotz
>
>Would there be padding between the Bar, Baz and Frotz arrays in this case?

Yes there would. The size of the union is the maximum of the sizes of its elements.
Suppose that the compiler padded everything to 32-bit boundaries. Then your
declaration is effectively:

union {
	struct foo {
	    char	Bar[3];
            char	pad1[1];
	    char	Baz[6];
            char	pad2[2];
	    char	Frotz[4];
	} fred;
	char	SomeString[3 + 6 + 4];
} Strings;

and the memory layout is:

     Strings+:    0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
     fred:        --Bar--  #  ------Baz------- ##### ---Frotz---
     SomeString:  0  1  2  3  4  5  6  7  8  9 10 11 12 ########

where # indicates memory via this union element.

The answer to the original posting is "No". The only way to do it is some
technique which declares a single array and then breaks it up (e.g. via
#defines).
-- 
Clive D.W. Feather           clive@ixi.uucp
IXI Limited                  ...!mcvax!ukc!acorn!ixi!clive (untested)
                             +44 223 462 131