From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-prs-return-17415-listarch-gcc-prs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 3057 invoked by alias); 17 Apr 2002 09:16:06 -0000
Mailing-List: contact gcc-prs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc-prs/>
List-Post: <mailto:gcc-prs@gcc.gnu.org>
List-Help: <mailto:gcc-prs-help@gcc.gnu.org>
Sender: gcc-prs-owner@gcc.gnu.org
Received: (qmail 3006 invoked by uid 71); 17 Apr 2002 09:16:02 -0000
Resent-Date: 17 Apr 2002 09:16:01 -0000
Resent-Message-ID: <20020417091601.3004.qmail@sources.redhat.com>
Resent-From: gcc-gnats@gcc.gnu.org (GNATS Filer)
Resent-To: nobody@gcc.gnu.org
Resent-Cc: gcc-prs@gcc.gnu.org, gcc-bugs@gcc.gnu.org
Resent-Reply-To: gcc-gnats@gcc.gnu.org, dh@digitalbrain.com
Received:(qmail 539 invoked by uid 61); 17 Apr 2002 09:11:09 -0000
Message-Id:<20020417091109.537.qmail@sources.redhat.com>
Date: Wed, 17 Apr 2002 02:16:00 -0000
From: dh@digitalbrain.com
Reply-To: dh@digitalbrain.com
To: gcc-gnats@gcc.gnu.org
X-Send-Pr-Version:gnatsweb-2.9.3 (1.1.1.1.2.31)
Subject: optimization/6330: tiny C++ prog broken under -O1
X-SW-Source: 2002-04/txt/msg00878.txt.bz2
List-Id: <gcc-prs.sourceware.org>


>Number:         6330
>Category:       optimization
>Synopsis:       tiny C++ prog broken under -O1
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Apr 17 02:16:00 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     dh@digitalbrain.com
>Release:        2.95.4 20011002 (Debian prerelease)
>Organization:
>Environment:
gcc version 2.95.4 20011002 (Debian prerelease)

Linux fee 2.4.18-p7-rmap12c-r #267 SMP Mon Feb 4 12:05:56 GMT 2002 i686 unknown
>Description:
Taking the address of an automatic variable says to the compiler "don't use a register for this!".

But I suspect that casting an automatic variable into another type and then taking the address of That
doesn't say anything of the kind
(even though it should)
>How-To-Repeat:
this linux console output shows a tiny shell script
being run that duplicates the problem ..

5750:fee:~: cat ./show_gcc_bug.sh
#!/bin/sh

# what compiler/OS version is it?
gcc -v
uname -a

# create a program
echo '#include <stdio.h>' > x.cpp
echo 'typedef unsigned char byte;' >> x.cpp;
echo 'void change(byte **z) { *z = (byte*)"correct"; }' >> x.cpp
echo 'int main() {' >> x.cpp
echo '    char *s = "incorrect";' >> x.cpp
echo '    change(&(byte*)s);' >> x.cpp
echo '    printf("%s\n", s);' >> x.cpp
echo '    exit(0);' >> x.cpp
echo '}' >> x.cpp

# compile and run it
gcc x.cpp
./a.out

# compile and run The Same Program optimized
gcc -O1 x.cpp
./a.out

5751:fee:~: ./show_gcc_bug.sh
Reading specs from /usr/lib/gcc-lib/i386-linux/2.95.4/specs
gcc version 2.95.4 20011002 (Debian prerelease)
Linux fee 2.4.18-p7-rmap12c-r #267 SMP Mon Feb 4 12:05:56 GMT 2002 i686 unknown
correct
incorrect
5752:fee:~:
>Fix:
no fix known ..

work arounds:

1)
change the x.cpp from &(byte*)s
to (byte**)&s

2)
rename x.cpp to x.c
because the C front end will give the error
"x.c:6: invalid lvalue in unary `&'"
rather than make incorrect code

>Release-Note:
>Audit-Trail:
>Unformatted: