• d-bugmail puremagic.com (49/49) Nov 30 2013 https://d.puremagic.com/issues/show_bug.cgi?id=11647

d-bugmail puremagic.com writes:

https://d.puremagic.com/issues/show_bug.cgi?id=11647

           Summary: Different D sequence point rules
           Product: D
           Version: D2
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: DMD
        AssignedTo: nobody puremagic.com
        ReportedBy: bearophile_hugs eml.cc



Wrong C code:


#include <stdio.h>
int main(){
    int i = 0;
    int a[] = {10,20,30};
    int r = 1 * a[i++] + 2 * a[i++] + 3 * a[i++];
    printf("%d\n", r); 
    return 0;
}


Clang gives on that code:

warning: multiple unsequenced modifications to 'i' [-Wunsequenced]


GCC 4.8.0 gives on that code:
test.c: In function 'main':
test.c:5:46: warning: operation on 'i' may be undefined [-Wsequence-point]
     int r = 1 * a[i++] + 2 * a[i++] + 3 * a[i++];
                                              ^
test.c:5:46: warning: operation on 'i' may be undefined [-Wsequence-point]


This similar D code compiles and prints "140":

void main() {
    import std.stdio;
    int i = 0;
    auto a = [10, 20, 30];
    int r = 1 * a[i++] + 2 * a[i++] + 3 * a[i++];
    r.writeln;
}


I suggest to statically forbid such D code, or to make it defined and
deterministic on all D compilers. (From my coding experience, I find that kind
of code often confusing for the programmer too, I refactor away that kind of
code, so probably I'd like it to be forbidden).


More info:
http://www.open-std.org/jtc1/sc22/wg14/www/docs/n925.htm
http://www.open-std.org/jtc1/sc22/wg14/www/docs/n926.htm
http://www.open-std.org/jtc1/sc22/wg14/www/docs/n927.htm

-- 
Configure issuemail: https://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------