NetNews Usenet Archive 1992 #20/NN_1992_20.iso/spool/comp/std/c/2584

Internet Message Format | 1992-09-08 | 1.7 KB

Path: sparky!uunet!mcsun!uknet!acorn!ixi!clive
From: [e-mail] (Clive Feather)
Newsgroups: comp.std.c
Subject: Re: void expressions and volatile types
Message-ID: <[e-mail]>
Date: 7 Sep 92 10:40:43 GMT
References: <[e-mail]>
Organization: IXI Limited, Cambridge, UK
Lines: 30

In article <[e-mail]> [e-mail] (T. William Wells) writes:
> The problem is then what to do with statements like a; and a=b;
> when a is a volatile object. The first expression accesses a but
> then discards the value, which is fine. The other does NOT do
> what you'd expect, though. The reason is that the assignment has a
> value which is the value of a after the assignment, which can only
> be obtained by accessing a. Thus a=b results in a sequence like:
>
>    load b
>    store a
>    load a
>
> rather than the intended load and store.

Here's the mistake. The value of the expression can be obtained by
remembering the value stored into a - it is not necessary to re-fetch
it, even though a is volatile.

The storing of a value into a is a side-effect, and so takes place
sometime before the next sequence point. Any other change that takes
place to a (e.g. because it is a memory-mapped device) between the same
pair of sequence points counts as a double assignment, and is therefore
undefined, and therefore the implementation can do what it wants
(including not reading a after storing).

--
Clive D.W. Feather     | IXI Limited         | If you lie to the compiler,
[e-mail]          | 62-74 Burleigh St.  | it will get its revenge.
Phone: [phone] | Cambridge   CB1 1OJ |   - Henry Spencer
Fax:   [phone] | United Kingdom      |