The following bug occurs only if -O (or higher optimization) is switched on.
I did not find out which of the single optimization switches is responsible for
it. The complete if statement is true, although none of each boolean value is
true. If I replace the enum variable "Type" by an int, the code works correct
even with optimization. Please let me know if this bug is already known or even
fixed. 


Compiler:
../configure --with-as=/usr/ccs/bin/as --with-ld=/usr/ccs/bin/ld --disable-nls
Thread model: posix
gcc version 3.4.1

System (uname -a)
SunOS sun2 5.9 Generic_112233-12 sun4u sparc SUNW,Sun-Blade-1500

//gcc -O   enumtest.cpp    (has the bug)
//gcc -O    enumtest.cpp  -lsupc++ (has the bug) 
//or: gcc    enumtest.cpp  -lsupc++ (is correct)
//output is(if used -O ): 
//>inside 
//>done


//output is(if not used -O ): 
//>done 
//(as I expect) 

#include <stdio.h>

enum ETestType
{
        e_Zero  = 0
,       e_One
,       e_Two
,       e_Three
,       e_Four
};


int main () {
        ETestType Type= e_Four;
        if( ( Type == e_One ) 
         || ( Type == e_Two    ) 
         || ( Type == e_Three   ) )
        {
                printf("inside\n");
        }
        printf("done\n");
        return 0;
}


David


-- 
           Summary: seems to be a bug in optimization on sparc systems.
           Product: gcc
           Version: 3.4.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: d dot obermann at callassoftware dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25014

Reply via email to