From: "XtReMe" <xtreme AT x-project DOT demon DOT nl>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: odd sizeof results
Date: Sun, 28 Feb 1999 02:22:21 +0100
Message-ID: <920164868.21105.0.rover.d4ee0fa5@news.demon.nl>
References: <7at84k$etf$1 AT news5 DOT svr DOT pol DOT co DOT uk> <36D23316 DOT 5506DAC8 AT a DOT crl DOT com>
NNTP-Posting-Host: x-project.demon.nl
X-NNTP-Posting-Host: x-project.demon.nl:212.238.15.165
X-Trace: news.demon.nl 920164868 rover:21105 NO-IDENT x-project.demon.nl:212.238.15.165
X-Complaints-To: abuse AT demon DOT net
X-Newsreader: Microsoft Outlook Express 4.72.3110.5
X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3
Lines: 46
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com


Weiqi Gao heeft geschreven in bericht <36D23316 DOT 5506DAC8 AT a DOT crl DOT com>...
>> "<---------------------------" wrote:
>>
>> I have a structure that contains an unsigned char (typedefed for
>> convenience) and a pointer to the next structure in my list.
>> Because list traversal has been giving some dodgy results I set
>> about checking everything - until I found this...
>>
>> typedef unsigned char number
>>
>> struct number_t
>> {
>>     number value;
>>     struct number_t *next;
>> };
>>
>> struct number_t test;
>>
>> sizeof(struct number_t) = 8 bytes???
>>
>> yet
>>
>> sizeof(test.value) = 1 and
>> sizeof(test.next) = 4.
>>
>> the last maths book I checked said that 1 + 4 was still five.
>> Any help would be greatly appreciated.
>
>The three extra bytes in the struct is caused by alignment and is
>normal.
>
Yes, but if you want those struct's to be as tiny as possible use the
'__attribute__ ((packed))' attribute. like this :

struct number_t
{
    number value __attribute__ ((packed));
    struct number_t *next __attribute__ ((packed));
};

But I still find it curious when you define these attributes, and the
struct's
still aren't packed. Has this to do with optimisations or something?