Newsgroups: comp.os.msdos.djgpp
From: design AT netcom DOT com ('Aaron Waters')
Subject: Re: sizeof(long double)==12
Message-ID: <designDxy9Ky.71q@netcom.com>
Organization: Design and Delivery
References: <51om9k$cof AT bagan DOT srce DOT hr>
Date: Wed, 18 Sep 1996 22:38:58 GMT
Lines: 23
Sender: design AT netcom5 DOT netcom DOT com
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp

In article <51om9k$cof AT bagan DOT srce DOT hr>, S. Mikecin <sime AT fly DOT cc DOT etf DOT hr> wrote:
>DJGPP has three floating-point (FP) data types: float (4 bytes), double (8
>bytes) and long double (12 bytes). Since x86 FPU supports ONLY
>32-bit,64-bit and 80-bit floating-point numbers (and also 64-bit integer
>numbers), I see no reason why long double is of size 12, and not 10 bytes!
>It has nothing to do with alignment because it is also 12, even when I use
>#pragma pack(1). Is there any explanation?

This is properly a gcc question, which belongs in the gnu newsgroups,
not a djgpp specific question.  One way to tell the difference is to ask
yourself:  is this issue I am asking about likely to be an issue under
Unix or OS/2 as well as under DOS?  If the answer is yes, you need to
talk to the gnu folks; if no, you have a true djgpp question.

That said, the answer is:  it has to do with alignment.  First of all,
last I checked, #pragma pack had no effect under gcc (RMS is strongly
opposed to any use of #pragma).  Second of all, even if you *did* use
the gnu equivalent (whatever that is these days--all I remember for sure
is that there is one), that only effects structures!  And a long double
is *NOT* a structure!  Not even in 16 bit WinDOS compilers!   ;)

cheers,
Chris