From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-117139-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 2324 invoked by alias); 12 Jul 2005 14:41:35 -0000
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc/>
List-Post: <mailto:gcc@gcc.gnu.org>
List-Help: <http://gcc.gnu.org/ml/>
Sender: gcc-owner@gcc.gnu.org
Received: (qmail 2073 invoked by uid 22791); 12 Jul 2005 14:41:24 -0000
Received: from nobodaddy.cs.utexas.edu (HELO nobodaddy.cs.utexas.edu) (128.83.120.154)
    by sourceware.org (qpsmtpd/0.30-dev) with ESMTP; Tue, 12 Jul 2005 14:41:24 +0000
Received: from charco.cs.utexas.edu (njn@charco.cs.utexas.edu [128.83.144.45])
	by nobodaddy.cs.utexas.edu (8.13.4/8.13.4) with ESMTP id j6CEfM4e013957
	for <gcc@gcc.gnu.org>; Tue, 12 Jul 2005 09:41:22 -0500 (CDT)
Received: (from njn@localhost)
	by charco.cs.utexas.edu (8.13.4/8.13.4/Submit) id j6CEfM6D025160;
	Tue, 12 Jul 2005 09:41:22 -0500
Date: Tue, 12 Jul 2005 14:41:00 -0000
From: Nicholas Nethercote <njn@cs.utexas.edu>
To: gcc@gcc.gnu.org
Subject: Some tests in gcc.c-torture rely on undefined behaviour?
Message-ID: <Pine.LNX.4.63.0507120914490.23582@charco.cs.utexas.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
X-SW-Source: 2005-07/txt/msg00503.txt.bz2

Hi,

I've been looking at the gcc.c-torture tests, it seems some of them rely 
on undefined behaviour.  For example, 920612-1.c looks like this:

   f(j)int j;{return++j>0;}
   main(){if(f((~0U)>>1))abort();exit(0);}

AIUI, this passes the largest possible positive integer to f(), which then 
increments it, causing signed overflow, the result of which is undefined. 
The test passes if signed overflow wraps.

930529-1.c is similar -- again the maximum positive integer is 
incremented.

20020508-2.c has the following code (abridged):

   #ifndef CHAR_BIT
   #define CHAR_BIT 8
   #endif

   #define ROR(a,b) (((a) >> (b)) | ((a) << ((sizeof (a) * CHAR_BIT) - (b))))

   #define INT_VALUE ((int)0x1234)

   #define SHIFT1 4

   int i = INT_VALUE;
   int shift1 = SHIFT1;

   if (ROR (i, shift1) != ROR (INT_VALUE, SHIFT1))
     abort ();

Similarly, the left-shifting in ROR causes signed integer overflow (I 
think) and so ROR relies on undefined behaviour.  20020508-3.c is similar.

My question is:  what exactly is gcc.c-torture testing?  It seems to be 
testing more than just C standard compliance, but also certain 
undefined-but-desired behaviours, such as wrap-on-signed-overflow (at 
least in some cases).  Is that right?  If so, other C compilers could be 
correct with respect to the C standard but not pass all the tests.  I 
couldn't find any kind of README or description about the tests that 
covered this point, so I'd appreciate any explanations.

Thanks.

Nick