Path: csiph.com!3.us.feeder.erje.net!feeder.erje.net!news.linkpendium.com!news.linkpendium.com!news.iecc.com!.POSTED.news.iecc.com!nerds-end
From: David Brown <david.brown@hesbynett.no>
Newsgroups: comp.compilers
Subject: Re: Bounds checking, Optimization techniques and undefined behavior
Date: Thu, 9 May 2019 09:19:23 +0200
Organization: A noiseless patient Spider
Lines: 59
Sender: news@iecc.com
Approved: comp.compilers@iecc.com
Message-ID: <19-05-074@comp.compilers>
References: <19-04-021@comp.compilers> <19-04-023@comp.compilers> <19-04-037@comp.compilers> <19-04-039@comp.compilers> <19-04-042@comp.compilers> <19-04-044@comp.compilers> <19-04-047@comp.compilers> <19-05-004@comp.compilers> <19-05-006@comp.compilers> <19-05-016@comp.compilers> <19-05-020@comp.compilers> <19-05-024@comp.compilers> <19-05-025@comp.compilers> <19-05-028@comp.compilers> <19-05-029@comp.compilers> <19-05-034@comp.compilers> <19-05-045@comp.compilers> <19-05-061@comp.compilers>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Injection-Info: gal.iecc.com; posting-host="news.iecc.com:2001:470:1f07:1126:0:676f:7373:6970"; logging-data="63993"; mail-complaints-to="abuse@iecc.com"
Keywords: standards, errors, comment
Posted-Date: 09 May 2019 11:28:50 EDT
X-submission-address: compilers@iecc.com
X-moderator-address: compilers-request@iecc.com
X-FAQ-and-archives: http://compilers.iecc.com
Content-Language: en-GB
Xref: csiph.com comp.compilers:2309

On 08/05/2019 10:03, David Brown wrote:

>
> [I'm looking at C11 and don't see where it says you can't cast away
> const.  What am I missing?  I agree any approach here is ugly and we're
> near the limits of what you can say in C. -John]
>

6.7.3p6 :

"""
If an attempt is made to modify an object defined with a const-qualified
type through use of an lvalue with non-const-qualified type, the
behavior is undefined.
"""


Consider this code:

int v = 10;
const int c = 20;

void change(const int * p, int n) {
	*(int*) p = n;
}

int foo(int a) {
	change(&v, a);		// OK
	change(&c, a);		// Usually compiles, but NOT OK

	return c;
}



You can cast away const (via pointers), but if the original declaration
of the object was const, it is undefined behaviour.

Compiling this with gcc x86 and optimisation (with
<https://godbolt.org>), you get this:

change:
        mov     DWORD PTR [rdi], esi
        ret
foo:
        mov     DWORD PTR v[rip], edi
        mov     eax, 20
        mov     DWORD PTR c[rip], edi
        ret
c:
        .long   20
v:
        .long   10


Although the call "change(&c, a)" attempts to modify "c" (which will
likely cause a fault on many systems), the compiler assumes the value of
"c" is still 20 for the return.
[Oh, look at that.  Tried it with clang, did the same thing returning a constant 20. -John]