NetNews Usenet Archive 1992 #18/NN_1992_18.iso/spool/comp/lang/c/12437

Internet Message Format | 1992-08-17 | 1.6 KB

Path: sparky!uunet!zaphod.mps.ohio-state.edu!caen!hellgate.utah.edu!dog.ee.lbl.gov!horse.ee.lbl.gov!torek
From: [e-mail] (Chris Torek)
Newsgroups: comp.lang.c
Subject: Re: Memory alignment of struct elements
Date: 17 Aug 1992 18:26:20 GMT
Organization: Lawrence Berkeley Laboratory, Berkeley
Lines: 22
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <adam.714060134@mcrware>
Reply-To: [e-mail] (Chris Torek)
NNTP-Posting-Host: 128.3.112.15

In article <adam.714060134@mcrware> [e-mail] (Adam Goldberg) writes:
>That's all well and good, but what if, say, you are coding an implementation
>of LIM EMS 4.0, which takes as a parameter a structure which, in assembler,
>looks like [radically machine-dependent example omitted].

>Is there any standard way to FORCE the C Struct to be identical to the
>Asm struct?

Well, obviously if you are talking about coding for compatibility with
a specific assembler interface, you are throwing portability off the
top of the Sears Tower (never to be seen again), so you buy the C
compiler that comes with the appropriate compilation directives---as
compile-time options or `__packed' keywords or #pragma's or whatever
---and use those.

To paraphrase a famous statement, ``If you want assembly language, you
know where to find it.''  It is no good asking for C to be portably
nonportable.  Some machines---like the ones I use---are physically
incapable of accessing unaligned data.  (It causes a machine fault.)
--
In-Real-Life: Chris Torek, Lawrence Berkeley Lab CSE/EE ([phone])
Berkeley, CA        Domain:    [e-mail]