https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85419
Bug ID: 85419
Summary: Incorrect determination of null pointer constant
Product: gcc
Version: 4.8.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: terra at gnome dot org
Target Milestone: ---
Created attachment 43949
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=43949&action=edit
Preprocessed source code
Priority: low
#include <stdio.h>
int
main (int argc,char **argv)
{
printf ("%p\n", 0 ? (void *restrict)0 : (int *)0);
return 0;
}
As I read C99's section 6.3.2.3, (void *restrict)0 is *not* a null pointer
constant.
I am reading section 6.5.15 as saying the ?: is valid with type
void *restrict.
%p is valid for such a pointer.
gcc disagrees:
# gcc -std=c99 -pedantic -Wall t.c
t.c: In function ‘main’:
t.c:6:3: warning: format ‘%p’ expects argument of type ‘void *’, but argument 2
has type ‘int *’ [-Wformat=]
printf ("%p\n", 0 ? (void *restrict)0 : (int *)0);
^
I.e., it thinks (void *restrict)0 is a null pointer constant. Hence ?:
becomes an int* and %p isn't valid for that in a pedantic sense.
Removing the "restrict" clearly makes the program invalid in a pedantic
sense. I.e., Section 6.7.3 #7 is wrong.
Details: stock OpenSuSE 42.3
