home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.std.c
- Path: vixen.cso.uiuc.edu!uchinews!news
- From: Michael Spertus<mps@geodesic.com>
- Subject: Re: Alignment of malloc()
- X-Nntp-Posting-Host: ford.uchicago.edu
- Message-ID: <DKKHCH.L6r@midway.uchicago.edu>
- Sender: news@midway.uchicago.edu (News Administrator)
- Organization: Dept. of Mathematics
- References: <DKDA7D.Kw7@midway.uchicago.edu> <j66Sx*FRe@yaps.rhein.de>
- Date: Tue, 2 Jan 1996 18:43:28 GMT
-
- In article <j66Sx*FRe@yaps.rhein.de>,
- Arno Eigenwillig <arno@yaps.rhein.de> wrote:
- >
- >That is a quality-of-implementation issue and hence beyond the scope
- >of The Standard. All The Standard defines in the end is the output a C
- >program creates, not the timing nor (thanks to the as-if rule) the way
- >it creates it.
- >
- My question is what the committee means by the words "proper alignment". If
- "proper" COMPLETELY ignores efficiency consideration even if they are
- gross, it means that I would have to align all malloc() returns according to
- the size of the data in all portable code I write. However, I've never seen
- an "ANSI C" textbook that advises aligning malloc() returns by hand. So
- the interpretation that completely ignores reasonable efficiency expectations
- is incompatible with the way people are taught to code portably in C.
-
- For this reason, the two possibilities are:
-
- 1) All the texts (including K & R) are advocating code that is not reasonable
- on (the most common) compliant implementations of C.
-
- or
-
- 2) 32 bit alignment of malloc() on PCs is not compliant.
-
- Either possibility is very serious, and violates assumptions that most
- programmers take for granted
-
-
- Regards,
-
- Mike
-